Vol.22no.92006,pages1080–1087
doi:10.1093/bioinformatics/btl046 BIOINFORMATICS ORIGINAL PAPER
Structural bioinformatics
Development and validation of a consistency based multiple structure alignment algorithm
Jessica Ebert and Douglas Brutlag?
Program in Biophysics and Department of Biochemistry,Stanford University,Stanford,CA94305USA
Received on November8,2005;revised on December20,2005;accepted on February6,2006
Advance Access publication February10,2006
Associate Editor:Thomas Lengauer
ABSTRACT
Summary:We introduce an algorithm that uses the information gained from simultaneous consideration of an entire group of related proteins to create multiple structure alignments(MSTAs).Consistency-based alignment(CBA)first harnesses the information contained within regions that are consistently aligned among a set of pairwise super-positions in order to realign pairs of proteins through both global and local refinement methods.It then constructs a multiple alignment that is maximally consistent with the improved pairwise alignments.We validate CBA’s alignments by assessing their accuracy in regions where at least two of the aligned structures contain the same conserved sequence motif.
Results:CBA correctly aligns well over90%of motif residues in super-positions of proteins belonging to the same family or superfamily,and it outperforms a number of previously reported MSTA algorithms. Availability:CBA is available at https://www.doczj.com/doc/d37417306.html,/and the source code is freely available at https://www.doczj.com/doc/d37417306.html,/software/ Contact:brutlag@https://www.doczj.com/doc/d37417306.html,
Supplementary information:Supplementary data are available at Bioinformatics online.
INTRODUCTION
Methods for the alignment of multiple proteins,whether guided by sequence,structure or both sources of information,are the corner-stone of many techniques in the biological sciences.A number of current areas of research depend on the ability to identify structur-ally equivalent residues across a group of proteins,including a variety of fold recognition and prediction techniques(Bystroff and Shao,2002),homology modeling(Madej et al.,1995; Panchenko et al.,1999),evolutionary studies of both protein families and entire organisms,and investigations into the relation-ships among structure,function and sequence(Thornton et al., 2000;Todd et al.,2001).As factors that contribute to errors and ambiguity often differ between sequence alignment and structure superposition algorithms(Marchler-Bauer et al.,2002),multiple structure alignment(MSTA)has a place nearly everywhere multiple sequence alignment(MSA)is used.In particular,since structure is often more conserved than sequence,structural alignment has the potential to be more accurate than sequence alignment below the so-called twilight zone of sequence similarity(Doolittle,1986). However,relatively few MSTA algorithms exist,and even fewer have been subjected to meaningful and clear validation.
Early MSTA efforts often focused on placing all structures in a common frame of reference by aligning each of them to a stationary reference,such as the structure that is in some sense the centroid of the group(Akutsu and Sim,1999).This approach,however,does little to take advantage of the fact that more information can be obtained from simultaneous consideration of all of the structures than from independent analyses of pairs of proteins.Similar methods instead iterate this process,in each phase aligning the structures to the previous iteration’s MSTA,typically by de?ning the consensus or average structure implied by the multiple align-ment(Gerstein and Levitt,1996;Taylor et al.,1994).Progressive algorithms align the most closely related structures?rst and even-tually incorporate the more distantly related ones,and some perform a?nal optimization step once all structures have been aligned (Russell and Barton,1992;Sali and Blundell;1990,Yang and Honig,2000).Since early iterations examine only a subset of the proteins,they may discard the optimal alignment as suboptimal. Alternatively,it is possible to use pairwise alignments as a starting point for building an MSTA and then determine which residues should be aligned using either structure-based scoring functions (Guda et al.,2004)or graph theoretic methods designed to?nd the best set of residue equivalencies across all structures that do not con?ict with one another(Sandelin,2005).Some recent algorithms make greater use of simultaneous consideration of all structures in the input set,typically by building up a multiple alignment out of aligned structural fragments(Dror et al.,2003; Shatsky et al.,2004)or using methods such as geometric hashing to identify structural equivalences across a group of proteins (Leibowitz et al.,2001).
MSTA algorithms that explicitly or implicitly rely on pairwise superpositions in some way depend on the principle of transitiv-ity,often referred to as consistency,among the pairwise align-ments(Gotoh,1990).Mathematically,the three possible pairwise alignments among three structures are said to be consistent if the residue registrations of two of the alignments predict that of the third.Though a set of pairwise alignments will be fully consistent with one another only in the simplest cases,consis-tency has successfully been used as a driving force in the multi-ple alignment of both sequences(Do et al.,2005)and structures (Ochagavia and Wodak,2004).In fact,one recent algorithm for obtaining the residue correspondences necessary for a multiple sequence or structure alignment is based entirely on the premise of‘relaxed transitivity’(Van Walle et al.,2003).This method constructs a graph whose nodes are residues and whose edges connect residues that are either aligned in a set of pairwise
?To whom correspondence should be addressed
1080óThe Author2006.Published by Oxford University Press.All rights reserved.For Permissions,please email:journals.permissions@https://www.doczj.com/doc/d37417306.html,
alignments or that,if aligned,would create transitive cycles in the graph.
Here,we present CBA(consisteny-based alignment),an MSTA algorithm that uses information from consistent and nearly con-sistent regions of pairwise alignments to create a MSTA.CBA focuses almost entirely on the problem of determining which resi-dues should be aligned given the relative orientations of the proteins once they have been superimposed.We use consistency?rst to?nd residue equivalencies across some or all structures in the input set and then to correct residue registration errors and gaps in the pair-wise alignments in order to add additional aligned residues to the MSTA.We also introduce changes to the LOCK2pairwise protein superposition algorithm(Shapiro and Brutlag,2004)that signi?c-antly enhance consistency among the alignments that are the starting point for CBA.CBA is a general framework for MSTA consisting of several modules that can be independently modi?ed or even replaced.Simple modi?cations would also adapt it to perform multiple sequence alignment.Because CBA never uses sequence information,its alignments are well-suited to the study of the relationships among sequence,structure and function.
To date,very few validation techniques have been proposed for MSTA or MSA methods.Alignment algorithms are often val-idated by comparison with databases containing multiple align-ments whose construction has been fully or partially guided by structure,such as HOMSTRAD(Mizuguchi et al.,1998), BaliBASE(Thompson et al.,1999)and OXBench(Raghava et al.,2003),but even alignments that are manually corrected or built entirely by hand are subject to error.Fully or semi-automated databases such as HOMSTRAD require explicit trust in the algorithms used to produce their alignments,thus establishing these methods as de facto gold standards.While the HOMSTRAD authors state that alignments are validated by hand,an incorrect automated alignment could easily prove to be a misleading starting point for manual analysis,particularly given the dif?culty of visu-alizing multiple structure https://www.doczj.com/doc/d37417306.html,parison of an algo-rithm with a database such as HOMSTRAD is therefore an implicit comparison with the alignment methods used to produce it,and thus lacks the generality we desire for validation of a new algorithm.Other attempts to quantify alignment quality through the use of simple measurements such as the alignment length and the distances between aligned alpha carbons lack substantive connections to alignment accuracy for anything other than closely related proteins.Though many complex scoring functions have been proposed,even a seemingly reasonable one is nonetheless arbitrary.
We instead validate CBA by examining alignments of protein structures containing the same conserved sequence motifs. Although global sequence alignments of proteins that share only limited sequence similarity are too prone to error to be used as gold standards for MSTA,distantly related proteins may nonetheless contain conserved local sequence motifs that are expected to align to one another.This method avoids the problem of ambiguous or incorrect global sequence alignments in regions of low sequence similarity(<20%)while still retaining the ability to assess align-ments of proteins with low overall sequence identity in a fully automated fashion.We apply this benchmark to several previously reported MSTA algorithms and demonstrate that CBA outperforms them for multiple alignments of structures from the same family or superfamily.METHODS
Validation datasets
We have aligned structures from each family and superfamily in release1.65 of the Structural Classi?cation of Proteins(SCOP)(Murzin et al.,1995)that contains at least three sequences with the same PROSITE pattern(Sigrist et al.,2002)or eMOTIF hit(Nevill-Manning et al.,1998).These structures come from an ASTRAL subset of the SCOP database created in such a way that no two sequences share>40%sequence identity(Brenner et al.,2000). Since not all PROSITE patterns have high speci?city,we excluded as too general for our purposes any pattern that recognizes structures from multiple SCOP folds.Only structures from SCOP’s all alpha,all beta,alpha+beta and alpha/beta classes were considered.
The478family validation sets have an average of6.3members,among which the average sequence identity as determined by BLAST(Altschul et al.,1990)is 16.7%,while the197superfamily validation sets have an average of13.4members whose pairwise sequence identities average14%. Hence,relatively few pairs of sequences in any given validation set reach the upper limit of40%sequence identity,and many pairs are within or below the twilight zone of sequence similarity.Among the family and superfamily datasets that contain at least two examples of the same eMOTIF,an average of4.0structures share the motif.Since more than one eMOTIF may be built to recognize the same functional region,slightly different versions of the same motif may cover different members of the validation dataset.On average,the number of structures containing any eMOTIF is4.6and5.7 for family and superfamily validation sets,respectively.The average number of structures containing the same PROSITE motif is2.6in the family val-idation datasets and4.7in the superfamily datasets.All datasets are available for download in text format at https://www.doczj.com/doc/d37417306.html,/distributions/CBA/ validationsets.html
LOCK2
Though the pairwise structural alignments CBA uses to build a multiple alignment may be created by any structural superposition algorithm,we use an improved version of LOCK2by default.LOCK2produces an initial superposition by aligning secondary structure elements as previously repor-ted(Shapiro and Brutlag,2004),and replaces the residue alignment phase developed by Singh and Brutlag(1997)with a dynamic programming algo-rithm that scores the alignment of a query and target residue pair according to both the distances between their beta carbons and also the angles between the vectors de?ned by the alpha and beta carbons(Fig.1).The use of the beta carbon allows LOCK2to encode a preference to align residues whose side chains point in the same general direction.Glycine residues’alpha hydro-gens are replaced with beta carbons,whose positions are determined using ideal bond lengths and angles.
Assessment of motif alignment accuracy
When two or more structures in a multiple alignment contain the same local sequence motif,we assess the accuracy of the alignment in the motif region using the sequence motif as a gold standard.We sum the number of motif residues correctly aligned over all pairs of structures containing the motif and normalize to the number of residue pairs examined to give the percent of motif residues correctly aligned.If two or more motifs overlap,we count each position only once.
ALGORITHM
Steps1–3:obtaining an initial multiple superposition Because pairwise structural alignment algorithms consider only two proteins at a time,they cannot use information regarding the full extent of the variability observed in a given fold to resolve ambi-guous regions of the alignment.Both imperfect scoring functions and this rather limited view of fold space lead to registration errors,
Consistency based multiple alignment
1081
and hence the pairwise alignments among a group of proteins tend to contain residue registrations that are not consistent with one another.This problem is exacerbated by the use of rigid body transformation methods,which may be unable to optimize the superposition in all structurally similar regions simultaneously.CBA seeks to resolve inconsistencies in alignments by considering an entire group of structures at once and extracting from them more information than is available from independent analysis of pairs of proteins.
The CBA algorithm consists of seven steps,each of which incor-porates additional information into the alignment (Fig.2).After computing the pairwise alignments among the proteins in the input set (Fig.2,Step 1),CBA places all of the structures in the input set into a common frame of reference (Fig.2,Step 2).This is particularly helpful in cases involving structures with repeated subdomains.Given three structures,two with multiple copies of the same subdomain and one with only one copy,LOCK 2might align the smaller structure to different repeats in the two larger structures,thus yielding completely inconsistent pairwise align-ments.While this is not necessarily incorrect,it is inconvenient for analyzing similarities across all three proteins.Following a number of progressive MSTA algorithms,CBA clusters all of the structures in the input set by average linkage using LOCK 2’s global alignment scores as measures of similarity.This creates a binary guide tree that we use to progressively trans-form all structures into the same frame of reference beginning at the bottom of the tree,where clusters contain the most similar struc-tures,and proceeding towards the root node.In the trivial case of a node at the bottom of the tree with only two descendants,we transform one child structure onto the other using the transformation matrix from the pairwise alignment and label the parent node with the resulting superposition.When a node has more than two des-cendants,we choose one descendant of the left child and one from the right such that the two selected structures have the highest alignment score among all such pairs,and then use their pairwise transformation to add all descendants from the left child to the right child’s MSTA.
Since this process may change some pairwise superpositions,we run LOCK 2’s residue alignment phase a second time to obtain improved pairwise residue correspondences (Fig.2,Step 3).Though LOCK 2’s residue alignment phase does employ an iterative algorithm in which each cycle uses the previous cycle’s residue registration to update the transformation of one protein onto the
other,these changes in the overlap between the query and target proteins are minor and serve only to make the correct residue registration more clear.CBA discards the changes to the trans-formations and retains only the new residue registration,thus maintaining the common frame of reference obtained in Step 2.From this point forward,the positions of the proteins in space remain ?xed.
Step 4:obtaining an initial residue registration across all structures
Although the third step of CBA places all proteins in the same frame of reference,this superposition does not directly imply an un-ambiguous residue registration across all of the structures in any but the simplest of alignment problems.For the sake of clarity,we make a distinction between the superposition of a group of proteins,which de?nes their relative positions in space,and their residue registration,which speci?es the groups of structurally equivalent residues that are aligned to one another.The superposition is ?xed from this point forward,and the remaining phases of the algorithm focus on the more dif?cult problem of determining the residue registration without using sequence information.
The fourth step of CBA applies a fast implementation of the Markov cluster algorithm (MCL)(Van Dongen,2000)to ?nd the residue registration across all of the proteins that is maximally consistent with the pairwise alignments obtained at the end of Step 3(Fig.2,Step 4).One may view this task as the problem of ?nding clusters in a graph whose nodes are residues and whose edges connect nodes aligned in the pairwise superpositions (Fig.3).Edge weights,which re?ect CBA’s con?dence that the residue correspondences from the pairwise alignments are correct,are proportional to the LOCK 2residue alignment scores described in the Methods section.If the pairwise alignments were perfectly consistent,this graph would consist of a series of fully connected clusters,such as the leftmost cluster of Figure 3,and it would have no edges connecting nodes in different clusters.
The MCL algorithm operates under the paradigm that a random walk beginning in a region of the graph that corresponds to a cluster—in this case,a column in the multiple alignment—will visit many of the nodes in that cluster before leaving it for another region.That is,regions of the proteins that are structurally equi-valent will give rise to groups of nodes in the graph that are con-nected to one another by many edges and that have relatively few edges connecting them to other regions of the graph.If the structural similarity in the region is strong,the edges connecting these nodes will furthermore be associated with large weights.
Formally,MCL maximizes the ?ow through the graph where it is already relatively strong and weakens it where it is already weak,thus strengthening correct residue correspondences,eliminating edges that describe incorrect correspondences,and adding edges between residues not aligned by LOCK 2when the principle of transitivity suggests that they are structurally equivalent.Each resulting cluster contains equivalent residues across a subset of the proteins in the input set and is the basis for a column in the multiple alignment.Since the MCL method does not enforce sequence order constraints and does not require that each cluster contain only one residue from each protein,CBA must deal with these matters in subsequent steps of the algorithm.In the meantime,one can consider the latter issue in particular as a re?ection of uncertainty in the residue
registration.
Fig.1.The LOCK 2residue alignment scoring function.LOCK 2scores the alignment of a query and target residue pair by considering both the distance between their beta carbons and the angle between the vectors defined by each protein’s C a and C b atoms.Beta carbon positions for glycine residues are determined using ideal bond lengths and angles.
J.Ebert and D.Brutlag
1082
Steps 5and 6:realignment of secondary structure elements
Even when large portions of two protein structures align well,individual helices and strands may superimpose poorly,thus making
identi?cation of residue correspondences dif?cult.Realigning each pair of superimposed secondary structure elements in isolation circumvents this limitation of rigid body superposition algorithms,which cannot always optimally superimpose all structurally similar regions of two proteins simultaneously (Fig.2,Step 5).Since more than one registration of two secondary structure elements of the same type is possible,CBA initially superimposes them onto one other using the residue correspondences identi?ed by the MCL algorithm in Step 4.This produces a new superposition that is fully consistent with the current MSTA but that is better suited to LOCK 2’s scoring functions than was the overlap of the two secondary structure elements in the global pairwise alignment.We run the LOCK 2residue alignment phase on the newly super-imposed pair of secondary structure elements to ?ll in gaps and to correct mistakes in the current multiple alignment’s residue registration.We then transfer the resulting residue correspondences to the global pairwise alignment,but discard the changes in the superposition of the two secondary structure elements so that the positions of the proteins in space remain unchanged.After repeating this process for all pairs of aligned secondary structure elements in all pairs of proteins,we run the MCL algorithm on the new pairwise alignments in order to update the MSTA’s residue correspondences (Fig.2,Step 6).
Step 7:realignment of local regions
The ?nal stage of the CBA algorithm seeks to identify and realign small regions of the multiple alignment that could not be resolved in any previous step (Fig.2,Step 7).Regions to be realigned are derived both from secondary structure elements with gaps and from areas containing MSTA columns established by the MCL algorithm that violate sequence order constraints or that contain more than one residue from the same protein.Each region contains one
target
Fig.2.The CBA
algorithm.
Fig.3.A graph theoretic approach to determining residue correspondences from pairwise alignments.CBA determines residue correspondences across a group of proteins by constructing a graph in which each node corresponds to a residue from one of the proteins to be aligned.Two nodes are connected by an edge if they are aligned in the pairwise alignment of the structures to which they belong.Thus,no edges exist between nodes of the same protein,and a given node from one protein is connected to at most one node from each additional protein.The leftmost column of nodes in the graph shown here is a connected component,and thus corresponds to a column in the multiple alignment.The two middle columns of nodes contain a registration error that links them together,and the lack of an edge between the nodes from structures three and four in the third column indicates that these residues were left unaligned in the corresponding pairwise alignment.The rightmost column of nodes reflects a deletion in structure 3.
Consistency based multiple alignment
1083
domain whose alignment is suspect,and each secondary structure element or loop may be marked for realignment more than once with different target domains.
CBA realigns the target domain’s residues within a particular region by scoring the placement of a residue into a column accord-ing to the number of pairwise alignments that agree with the assign-ment.The assessment of structural similarity is therefore not based on an isolated and possibly misleading local perspective,but is instead derived solely from the global pairwise alignments obtained at the end of the sixth step of CBA.As was the case up to this point in the algorithm,sequence similarity is not considered.Dynamic programming then determines the new registration of the target domain in this region.This process thus uses the consistency among the pairwise alignments to correct local regions of the mul-tiple alignment.After realigning all regions marked as potentially incorrect,we iterate until no further changes are made or for a maximum of10iterations.
RESULTS
Within the391superfamilies in SCOP version1.67containing at least three members in an ASTRAL subset of domains constructed such that no two share>25%sequence identity,LOCK2’s pairwise residue registrations are on average68.2%consistent.This number is calculated from every triple of aligned residues from each triple of structures within the same superfamily.This high level of con-sistency among pairwise alignments often allows CBA to align distantly related protein structures.The average consistency within a superfamily increases to80.4%after obtaining a common frame of reference and recalculating the residue registration in the implied pairwise alignments in Steps2and3.After the realignment of secondary structure elements in Step5,the average consistency reaches84.2%.In the351families in SCOP containing at least three members in the25%sequence identity ASTRAL subset, the consistency among the initial LOCK2pairwise alignments averages76.7%.This value increases to85.1and87.7%after Steps3and5,respectively.These increases in consistency occur without a decrease in the number of aligned residues.
Because CBA observes sequence order constraints,we can read off a sequence alignment from the structural superposition. Figure4a shows a portion of an alignment of the17structures in SCOP’s globin-like superfamily that are included in the ASTRAL-25subset;members of four families,including two
phycocyanin domains,are included in the superposition.The full alignment(see Supplementary information)required2.2min.of CPU time(https://www.doczj.com/doc/d37417306.html,er time,42s system time)on a3.06Ghz Intel Xeon processor.Though gaps do occasionally appear in sec-ondary structure elements,many are justi?ed structurally or occur in regions where the correct residue alignment might be considered ambiguous.For example,the superposition of two of the structures, SCOP domains d1itha_and d1b0b__,shows the insertion of a residue in SCOP domain d1itha_(asparagine17)that induces a gap in globin helix A(Fig.4b).The alignment of d1itha_and d1b0b__shown in Figure4b suggests that it is reasonable to leave either16L or17N of d1itha_unaligned.In cases such as this,CBA tends to align the residue that is most consistently aligned in the pairwise alignments.Many other gaps in helices and strands are induced by insertions of loop residues that do not align to the secondary structure element.Validation by comparison to sequence motifs
Since manual analysis and validation of multiple structural align-ments can be both prone to error and too time consuming to approach on a large scale,we sought to validate CBA by examining its alignments in regions containing conserved sequence motifs. Even when a set of proteins is so diverse that sequence alignment methods are not reliable enough to be used as a gold standard,a subset of the sequences may still contain the same local sequence motif.If the motifs themselves are reliable,then the alignment accuracy can be assessed in these regions by assuming that residues at the same position of the motif should align to one another.We use both PROSITE patterns(Sigrist et al.,2002),which are constructed manually to represent known functional regions,and eMOTIFs (Nevill-Manning et al.,1998),which are built automatically from high quality multiple alignments of proteins sharing
strong Fig.4.Structurally reasonable gaps in a superposition of proteins from four globin-like families(a)A portion of the CBA alignment of the17members of SCOP’s globin-like superfamily contained in an ASTRAL-40subset is shown.The first two structures are protozoan/bacterial hemoglobins (whose A helices are truncated or deleted),the third is a nerve tissue mini-hemoglobin,the next12are globins from various species,and the last two are phycocyanins.Helical residues are shaded.The arrow indicates a residue(17N of d1itha_)that induces a gap in helix A.(b)The superposition of SCOP domains d1itha_and d1b0b__,whose relative orientation with respect to d1itha_is typical of other aligned domains,reveals that the gap induced by the insertion is structurally reasonable.
J.Ebert and D.Brutlag 1084
sequence similarity.Since CBA never uses sequence information, this is an independent test of its accuracy.
In an ASTRAL subset of SCOP(version1.65)created such that no two structures have>40%sequence identity,there are 478families and153superfamilies in which at least two structures contain the same PROSITE pattern.Similarly,there are123families and121superfamilies in which at least two structures contain the same eMOTIF.All members of the SCOP family or superfamily represented in the ASTRAL-40database are included regardless
of whether they contain the sequence motif;this allows for a more dif?cult test by increasing both the number of structures to be superimposed and the overall structural variability.The average pairwise sequence identity as determined by BLAST within these validation sets is 17%for families and14%for superfamilies. For each pair of proteins in a validation set that contain a given sequence motif,we count the number of motif positions that CBA correctly aligned to produce a sum of pairs measurement of align-ment accuracy.When a motif occurs in more than one location in one or both proteins,we consider only the pair of hits that is best aligned according to this criterion.Summing over all pairs of proteins with the motif yields a measure of the accuracy of the CBA alignments in these regions.CBA correctly aligns96.4% of eMOTIF residues and97.4%of PROSITE residues in the family validation sets(Table1).Interestingly,only92%of eMOTIF residues were correctly aligned in the LOCK2pairwise superpositions used by CBA.Hence,CBA was able to take advant-age of the information contained in a set of related structures in order to increase alignment accuracy.In the signi?cantly more challenging superfamily validation sets,CBA correctly aligns 91.6%of eMOTIF residues and92.6%of PROSITE residues. When an alignment of135P-loop containing nucleoside triphos-phate hydrolases is excluded from this analysis,these numbers increase to94.9and94.2%,respectively.This procedure examines 17922aligned residue pairs in eMOTIF hits and12783pairs in PROSITE patterns within the family validation sets,though there is a great deal of overlap between these two databases.The corres-ponding numbers for the superfamily validation sets are18485and 14647,respectively.
Comparison with CEMC,MultiProt and MASS
The use of sequence motifs as a gold standard for alignment accuracy provides an objective benchmark for evaluating MSTA algorithms.Since the accuracy of the alignment in regions contain-ing sequence motifs was assessed only after algorithmic develop-ment had concluded,CBA was not overtrained to optimize the alignment of motifs and hence does not have an inherent advantage from this perspective over other algorithms.
We benchmarked CEMC(Guda et al.,2004),MultiProt(Shatsky et al.,2004)and MASS(Dror et al.,2003)using the same sets of multiple alignment validation sets as reported above for CBA.
A minor modi?cation of the CEMC source code was necessary to prevent it from excluding structures that it considered to be too dissimilar to superimpose simultaneously.In some cases,CEMC or MASS did not produce a superposition at all.Because these failures may be more the result of technical issues than of algorithmic inadequacies,we removed these alignments from the validation sets.
Table2reports the percent of motif residues correctly aligned for all four multiple alignment algorithms using the sum of pairs score described above for assessing the alignment accuracy in regions containing sequence motifs.Both MASS and MultiProt report more than one possible multiple superposition for a given set of structures,so we count only the solution that yields the highest alignment accuracy with respect to the sequence motifs.This approach is rather lenient,as it does not require MASS and MultiProt to identify the best alignment before the motif alignment accuracy is assessed.Because MASS and MultiProt do not always produce a solution that superimposes all domains,we used two different methods to score their alignments.In the?rst method [Table2,method(a)],we consider only alignments that include all domains,and record a score of zero for cases in which the algorithm did not produce a solution meeting this criterion. Since MASS frequently fails to align all structures in the superfam-ily validation sets,we evaluated it using this scoring method only for the family validation sets.
Though structures that do not contain the motif under consideration were deliberately added to the validation sets to increase their dif?culty,we also computed scores for MASS and MultiProt without requiring them to superimpose all domains [Table2,method(b)].While this clearly gives MASS and MultiProt a distinct advantage over algorithms that superimpose all structures, they are still outperformed by CBA and CEMC.The average percentage of structures that were aligned in the best performing superposition is given in parentheses in the columns corresponding to method(b)in Table2.
Table1.Percent of motif residues correctly aligned by CBA
Family validation Sets(%)Superfamily
validation
Sets(%)
Reduced
set of
superfamilies a(%)
eMOTIFs96.491.694.9
PROSITE patterns97.492.694.2
a Results without considering the P-loop containing nucleoside triphosphate hydrolase https://www.doczj.com/doc/d37417306.html,parison of motif alignment accuracies
CBA CEMC MultiProt MASS
Method
(a)(%)
Method
(a)(%)
Method
(a)(%)
Method
(b)(%)
Method
(a)(%)
Method
(b)(%)
eMOTIF
families
96.293.475.187.7(77.5)65.180.2(73.2)
eMOTIF
superfamilies
94.887.747.384.9(51.3)N/A61.9(54.3) PROSITE
families
97.595.981.790.7(72.9)72.383.2(73.1)
PROSITE
superfamilies
95.489.455.387.8(57.3)N/A60.1(53.2)
Method(a):Only superpositions that align all domains are considered.A score of zero is given when an algorithm does not produce an alignment that superimposes all domains. Method(b):Superpositions that do not align all domains are permitted.The average percent of structures superimposed in the multiple alignments that achieve the greatest motif alignment accuracies is given in parentheses.
Consistency based multiple alignment
1085
CBA reaches higher accuracies than the other three algorithms for both the family and superfamily validation sets regardless of the scoring method used,though CEMC does approach CBA’s accur-acy in the family validation tests.The most striking improvement over the previously reported MSTA methods,however,occurs in the superfamily validation sets.Even when MultiProt and MASS are permitted to exclude an arbitrary number of structures from the alignment[scoring method(b)],CBA’s ability to align distantly related structures is clearly superior.
DISCUSSION
CBA provides a general framework for building multiple structure superpositions from pairwise alignments,and it is easily modi?ed for sequence alignment problems by removing steps such as the realignment of individual secondary structure elements that may not be applicable in this case.We address the problem of resolving inconsistencies among pairwise alignments both by?nding a common frame of reference for all of the structures and by globally and locally re?ning the multiple alignment’s residue registration. CBA’s output consists of the superposition of the proteins and their residue registration.
Though CBA uses LOCK2by default to perform pairwise align-ments,any alignment algorithm may be substituted.The MCL clustering algorithm,which CBA uses several times to identify residue correspondences across multiple proteins,can similarly be replaced by another clustering algorithm,though we strongly caution against using progressive methods since they tend to propagate mistakes that occur in stages that consider only a subset of the proteins.Finally,the scoring system used in the last stage of CBA to realign small regions of the multiple-sequence alignment implied by the MSTA can be modi?ed,though we have found in practice that more complex scoring functions based on structural measurements increase the algorithm’s computational complexity without improving the resulting alignments(data not shown).CBA is designed to use no sequence information in producing a multiple alignment from pairwise alignments,but amino acid substitution scores would be an appropriate addition to the realignment scoring function in the case of MSA.
Assessment of multiple structure alignment’s accuracy in regions containing conserved sequence motifs represents a new method for multiple alignment validation.Though it can be argued that comparison with hand-curated,full length sequence or structure alignments is a more thorough approach,restricting the comparison to regions containing conserved sequence motifs provides a more straightforward analysis since this process is less subject to error than is the comparison with global sequence or structural align-ments of distantly related proteins.BaliBASE,a database of sequence alignments that are manually corrected using evidence from structural superpositions,recommends that comparisons be restricted to reliably aligned core blocks.Our approach is similar, but the use of sequence motifs is completely automated and does not rely on the ability to construct accurate,global alignments of diverse sequences or structures.Since BaliBASE alignments often contain sequences for which no structures exist,we did not perform a full-scale comparison with CBA.
Inclusion of structures in our validation sets that do not contain the sequence motif under consideration allows for tests of alignment accuracy at varying levels of dif?culty with respect to sequence identity,structural similarity and the number of structures in the validation set.Though the alignment accuracy obviously cannot be assessed in structures that do not contain the motif,their presence makes the multiple alignment task more dif?cult.Despite the fact that we have constructed validation sets from an ASTRAL subset of SCOP domains in which structures may have up to40%sequence identity,the average pairwise identity within a validation set is in practice<20%.Both the sequence similarity and the structural similarity of the proteins in the validation sets can easily be tuned for different applications.
CBA’s highly accurate alignment of the sequence motifs contained within SCOP domains allows us to evaluate the accuracy of the sequence motifs themselves.We?nd,for example,that a number of misaligned or unaligned eMOTIF residues occur at or near the ends of the motifs,suggesting that the automated procedure used to generate them may occasionally extend a motif beyond its true boundaries with respect to its functional or structural role. Pruning the motifs in regions of poor structural conservation may increase the sensitivity of eMOTIFs,or of motifs from any other source,without decreasing their speci?city.
Since PROSITE patterns are constructed manually,we observe misaligned residues at the ends of motifs more rarely than was the case for eMOTIFs.Instead,we have found instances in which a member of a validation set that only partially matched the PROSITE pattern nonetheless superimposed well in the region of the motif. Here,the PROSITE pattern may be too restrictive.Given a family of protein structures,several of which contain the same PROSITE pattern,we can again increase the sensitivity of the pattern without decreasing its speci?city by expanding the set of allowed amino acids at one or more positions in order to accommodate the members of the family that superimpose well within the motif region and nearly match the original sequence pattern.Structural information has been added to PROSITE patterns in the past in order to compensate for‘softened’substitution groups that are less restrictive than in the original pattern,but previous methods required a structural comparison between the PROSITE pattern and the target protein and hence were only useful for comparison to proteins of known structure(Jonassen et al.,2000).
For example,only two of the three structures in our annexin validation set match the annexin PROSITE pattern,but the third structure(chain A of the?rst domain of human annexin I)matches the sequence pattern at all but the last of its53positions.Its sequence contains a tryptophan at this unmatched position rather than one of the several hydrophobic amino acids speci?ed by the PROSITE pattern.Since CBA accurately aligned this position to the other two structures in the validation set,we created a modi?ed annexin pattern by adding a tryptophan to the last position’s substitution group.This new pattern picks up two additional matches in SCOP,both of which occur in a plant annexin(1dk5) and have a tryptophan residue in the last position,without encountering any additional false positives.The use of structural information compensates for the risk that increasing the number of amino acids permitted at a given position will decrease the PROSITE pattern’s speci?city.
Comparison of CBA with CEMC,MultiProt and MASS indi-cates that CBA represents a substantial improvement over these previously reported algorithms,particularly in the case of more distantly related structures.CEMC’s results in the case of proteins belonging to the same family lag behind CBA only by a relatively
J.Ebert and D.Brutlag 1086
small margin,but CBA clearly outperforms all three algorithms in the superfamily validation sets. ACKNOWLEDGEMENTS
This work was supported by NIGMS training grant number GM63495.
Conflict of Interest:none declared.
REFERENCES
Akutsu,T.and Sim,K.L.(1999)Protein threading based on multiple protein structure alignment.Genome Inform.Ser.Workshop Genome Inform.,10, 3–12.
Altschul,S.F.et al.(1990)Basic local alignment search tool.J.Mol.Biol.,215, 403–410.
Brenner,S.E.et al.(2000)The astral compendium for protein structure and sequence analysis.Nucleic Acids Res.,28,254–256.
Bystroff,C.and Shao,Y.(2002)Fully automated ab initio protein structure prediction using i-sites,hmmstr and rosetta.Bioinformatics,18(Suppl1),S54–S61.
Do,C.et al.(2005)Probcons:probabilistic consistency-based multiple sequence alignment.Genome Res.,15,330–340.
Doolittle,R.(1986)Of urfs and orfs:A Primer on How to Analyze Derived Amino Acid Sequences.University Science Books,Mill Valley,California.
Dror,O.et al.(2003)Mass:multiple structural alignment by secondary structures.
Bioinformatics,19(Suppl1),i95–i104.
Gerstein,M.and Levitt,M.(1996)Using iterative dynamic programming to obtain accurate pairwise and multiple alignments of protein structures.Proc.Int.Conf.
Intell.Syst.Mol.Biol.,4,59–67.
Gotoh,O.(1990)Consistency of optimal sequence alignments.Bull.Math.Biol.,52, 509–525.
Guda,C.et al.(2004)Ce-mc:A multiple protein structure alignment server.Nucleic Acids Res.,32,W100–W103.
Jonassen,I.et al.(2000)Searching the protein structure databank with weak sequence patterns and structural constraints.J.Mol.Biol.,304,599–619.
Leibowitz,N.et al.(2001)Musta—a general,ef?cient,automated method for multiple structure alignment and detection of common motifs:Application to proteins.
https://www.doczj.com/doc/d37417306.html,put.Biol.,8,93–121.
Madej,T.et al.(1995)Threading a database of protein cores.Proteins,23, 356–369.
Marchler-Bauer,A.et al.(2002)Comparison of sequence and structure alignments for protein domains.Proteins,48,439–446.
Mizuguchi,K.et al.(1998)Homstrad:A database of protein structure alignments for homologous families.Protein Sci.,7,2469–2471.Murzin,A.G.et al.(1995)Scop:A structural classi?cation of proteins database for the investigation of sequences and structures.J.Mol.Biol.,247,536–540.
Nevill-Manning,C.G.et al.(1998)Highly speci?c protein sequence motifs for genome analysis.Proc.Natl https://www.doczj.com/doc/d37417306.html,A,95,5865–5871.
Ochagavia,M.E.and Wodak,S.(2004)Progressive combinatorial algorithm for multiple structural alignments:application to distantly related proteins.Proteins, 55,436–454.
Panchenko,A.et al.(1999)Threading with explicit models for evolutionary conser-vation of structure and sequence.Proteins,37,133–140.
Raghava,G.P.et al.(2003)Oxbench:A benchmark for evaluation of protein multiple sequence alignment accuracy.BMC Bioinformatics,4,47.
Russell,R.B.and Barton,G.J.(1992)Multiple protein sequence alignment from tertiary structure comparison:Assignment of global and residue con?dence levels.
Proteins,14,309–323.
Sali,A.and Blundell,T.L.(1990)De?nition of general topological equivalence in protein structures.A procedure involving comparison of properties and relation-ships through simulated annealing and dynamic programming.J.Mol.Biol.,212, 403–428.
Sandelin,E.(2005)Extracting multiple structural alignments from pairwise alignments:A comparison of a rigorous and a heuristic approach.Bioinformatics, 21,1002–1009.
Shapiro,J.and Brutlag,D.(2004)Foldminer:Structural motif discovery using an improved superposition algorithm.Protein Sci.,13,278–294.
Shatsky,M.et al.(2004)A method for simultaneous alignment of multiple protein structures.Proteins,56,143–156.
Sigrist,C.J.et al.(2002)Prosite:A documented database using patterns and pro?les as motif descriptors.Brief Bioinform,3,265–274.
Singh,A.P.and Brutlag,D.L.(1997)Hierarchical protein structure superposition using both secondary structure and atomic representations.Proc.Int.Conf.Intell.Syst.
Mol.Biol.,5,284–293.
Taylor,W.R.et al.(1994)Multiple protein structure alignment.Protein Sci.,3, 1858–1870.
Thompson,J.D.et al.(1999)Balibase:A benchmark alignment database for the evalu-ation of multiple alignment programs.Bioinformatics,15,87–88.
Thornton,J.M.et al.(2000)From structure to function:Approaches and limitations.
Nat.Struct.Biol.,7,991–994.
Todd,A.E.et al.(2001)Evolution of function in protein superfamilies,from a structural perspective.J.Mol.Biol.,307,1113–1143.
Van Dongen,S.(2000)Graph Clustering by Flow Simulation.University of Utrecht, Amsterdam.
Van Walle,I.et al.(2003)Consistency matrices:quanti?ed structure alignments for sets of related proteins.Proteins,51,1–9.
Yang,A.S.and Honig,B.(2000)An integrated approach to the analysis and modeling of protein sequences and structures.Iii.A comparative study of sequence conservation in protein structural families using multiple structural alignments.J.Mol.Biol., 301,691–711.
Consistency based multiple alignment
1087
一、 二、 三、 填空题（每小题 分，共 ?分） 、设原??问题为?????? ?≥-=++-≥--≤++++-= ,0,5232 4 7 532min 3213213213213 21无约束x x x x x x x x x x x x x x x Z 则它的标准形和对偶规划问题分别为：________________________ 和 ________________________。 、用分枝定界法求整数规划12 12121121min 5 2 56 30 4,0Z x x x x x x x x x x =---≥-??+≤?? ≤??≥?且为整数 的解时，求得放松问题的解为? ＝ ? ? ? ? ? ?，则可将原问题分成如下两个子问题 与 求解。 、右图的最小支撑图是。 、右边的网络图是标号算法中的图，其中每条弧上的数 表示其容量和流量。该图中得到的可行流的增广链 (-3,1) (2,1) ②5(4) ④ ① 6(6) 6(4) ⑥ (0, ∞) 8(8) 3(2 ) 9(9)(5,1)
为： ，在其上可增的最大流量 为 。 、已知某线性规划问题，最优单纯形表如下 则其最优解为： ，最优值 max Z 。 二、单项选择题（每小题 分，共 分） 、下列表格是对偶单纯形表的是（ ? ）
、关于线性规划模型的可行域，叙述正确的为（ ） ?、可行域必有界； 、可行域必然包括原点； 、可行域必是凸的； 、可行域内必有无穷多个点。 、在运输问题中如果总需求量大于总供应量，则求解时应（ ） ?、虚设一些供应量； ?、虚设一个供应点； 、根据需求短缺量，虚设多个需求点； ?、虚设一个需求点。 、下列规划问题不可用动态规划方法求解的是（ ） ?、背包问题； ?、最短路径问题 、线性规化： ???≥≥=++++=0 ,010 34..max 321 3 32211y x x x x t s x c x c x c Z ?、22 min (,)(2)3(1).. 460,0f x y x y s t xy y x y ?=++-?+
精读课文和略读课文的异同 双柏妥甸小学：鲁先丽 摘要：新课标指出：各个学段的阅读教学都要重视朗读和默读。加强对阅读的指导，让学生逐步学会精读，略读和浏览，可见略读只是一种与精读和浏览并列的阅读方法。而略读课文作为课文的一种形式呈现在师生面前时，它只是个例子，一个提供教师教与学生学习略读的载体。略读课文中并不排斥精读，因此我认为略读课文中可以略中显精，培养学生掌握一些学习课文的方法。 略读课文的教学，最关键的是教师自身对文本的认真解读和准确把握，要使课堂在思路清晰、条理简约的基础上不失丰赡，应是“教”略而“学”丰。这个“略”度应该介于精读与独立阅读之间，不是囫囵吞枣，更不是越俎代庖牵着学生走，同时在学生自读感悟的基础上又不失教师有效的引导，让学生沉浸在文本之中，和文本一起呼吸、一起慨叹、一起喜怒哀乐，在实践中掌握读书方法，逐步提高独立阅读的能力，促进学生语文素养的形成和发展，这其实是快乐的心灵之旅，“精略相辅”的幸福之旅！长久以来，略读课文在语文教学中处于比较尴尬的境地：一方面，它是课文中的组成部分，不教不行；另一方面，正因为它是“略读”课文，因此，在教学中往往被大多教师所忽视，在实际教学中，教师往往轻描淡写，一读而过，学生对课文也没留下多大印象。长此以往，限制了学生的阅读能力发展。对此，我们必须重新审视略读课文教学，采取有效的教学策略，使略读课文教学达到它应有的目标。以下是我在略读课文教学过程中的一些想法和采取的策略。 策略一：紧抓提示，凸显重点 我们知道略读课文一般只有一课时，在这短短的一课时中，教师快速有效地抓准略读课文的教学重点，使略读课文和精读课文形成一个系统，形成教学合力，更好地发挥训练阅读，迁移能力和陶冶情操的功能。如何抓？从哪里抓？每篇略读课文前的导语为我们提供了很好的借鉴。人教版实验教材在每一课略读课文前都有一段承上启下的提示语从中体会到了什么。有条件的还可以搜集有关丝绸之路的故事和同学交流。提示语优美生动，既激发了学生的阅读兴趣，延续了本组课文的主题，又明确提出了这篇课文的阅读要求。我们可以把这段提示语概括为两个方面： 一、读课文，了解课文讲了哪些内容，从中体会到了什么。 这是要求我们在教学初始引导学生整体把握课文内容，了解文中介绍的主要内容和历史意义，激发学生热爱祖国西部的情感。 二、搜集有关丝绸这路的故事和同学交流。 这点则是要求我们在学完课文的同时，能够引导学生课后去搜集资料，培养学生搜集信息的能力，同时也丰富学生的课外知识，增进与同学间的合作。 像这样，对提示语进行二次解读，既有助于我们教师找准重点，又能及时有效地展开略读课文的教学。 策略二：注重预习，培养自主 苏霍姆林斯基曾经说过这样一段精辟的论语：“在人的心灵深处，都有一种根深蒂固的需要，这就是希望自己是一个发现者、研究者、探索者。在儿童的精神世界里，这种需要特别强烈。”在教学中，教师应根据学生的这一心理特点，激
《管理运筹学》期中测试题 第一部分 线性规划 一、填空题 1．线性规划问题是求一个 目标函数 在一组 约束条件 下的最值问题。 2．图解法适用于含有 两个 _ 变量的线性规划问题。 3．线性规划问题的可行解是指满足 所有约束条件_ 的解。 4．在线性规划问题的基本解中，所有的非基变量等于 零 。 5．在线性规划问题中，基本可行解的非零分量所对应的列向量线性 无 关 6．若线性规划问题有最优解，则最优解一定可以在可行域的 顶点_ 达到。 7．若线性规划问题有可行解，则 一定 _ 有基本可行解。 8．如果线性规划问题存在目标函数为有限值的最优解，求解时只需在其 可行解 的集合中进行搜索即可得到最优解。 9．满足 非负 _ 条件的基本解称为基本可行解。 10．在将线性规划问题的一般形式转化为标准形式时，引入的松驰变量在目标函数中的系数为 正 。 11．将线性规划模型化成标准形式时，“≤”的约束条件要在不等式左_端加入 松弛 _ 变量。 12．线性规划模型包括 决策变量 、目标函数 、约束条件 三个要素。 13．线性规划问题可分为目标函数求 最大 _ 值和 最小 _值两类。 14．线性规划问题的标准形式中，约束条件取 等 _ 式，目标函数求 最大 _值，而所有决策变量必须 非负 。 15．线性规划问题的基本可行解与基本解的关系是 基本可行解一定是基本解，反之不然 16．在用图解法求解线性规划问题时，如果取得最值的等值线与可行域的一段边界重合，则 _ 最优解不唯一 。 17．求解线性规划问题可能的结果有 唯一最优解，无穷多最优解，无界解，无可行解 。 18.如果某个约束条件是“ ”情形，若化为标准形式，需要引入一个 剩余 _ 变量。 19.如果某个变量X j 为自由变量，则应引进两个非负变量X j ′ , X j 〞, 同时令X j ＝ X j ′ - X j 〞 j 。 20.表达线性规划的简式中目标函数为 线性函数 _ 。 21.线性规划一般表达式中，a ij 表示该元素位置在约束条件的 第i 个不等式的第j 个决策变量的系数 。 22．线性规划的代数解法主要利用了代数消去法的原理，实现_ 基变量 的转换，寻找最优解。 23．对于目标函数最大值型的线性规划问题，用单纯型法代数形式求解时，当非基变量检验数_ 非正 时，当前解为最优解。 24．在单纯形迭代中，选出基变量时应遵循_ 最小比值 法则。 二、单选题 1． 如果一个线性规划问题有n 个变量，m 个约束方程(m 关于精读课文和略读课文的区别 在小学语文教材中，精读课文的编排占据主体地位。就其涵盖的内容而言，包含了字词、句、段、篇等语文基本功训练项目，范围十分广泛，知识含量极高，是进行语文基本功训练的主要凭借。由此可见，精读课文的教学在小学语文教学中处于重中之重的地位，它是完成小学语文教学任务的重要途径。但是，就篇幅而言，略读课文在小学中段所占比例较小，但到了高段则是二分之一。因此略读课文的教学同样不可忽视。下面我就精读课文和略读课文教学谈谈区别。 精读。特级教师林润生是这样说的："精读课文的教学，应让学生拥有完整的阅读过程，经历精读与略读，朗读与默读，自悟与交流，思考与想象，课文阅读与资料收集利用，感知内容与领悟感情，理解与积累运用语言的过程。以此让学生经历了完整的阅读过程，从而激发学生对大自然的热爱。.通过品词、析句，感情朗读，想象画面，让学生感悟语言生动形象和描写的优美。 略读课更需整体把握，不宜肢解课文，应采取读几遍的方法。如，采用自己喜欢的读书方式一读，读对读通，口诵新惟；二读，画画批批，粗知大意；三读，交流收获（可以是受到教育启发的，可以是好词佳句的，可以是文章写法的，可以是读书方法的），积累语言。 略读教学不等于略学。略读课文就是让孩子在精读课文中学到的方法运用到自己的学习中，并在读中掌握读书方法，提高阅读能力，我们要敢于放手，不要将课文"嚼"得过烂再"喂"给学生，教师的作用就在于适当点拨和引导。这就对教师提出了更严的要求，我们要注意讲读与自读的知识迁移与能力训练的衔接，完成"教"向"学"的过渡转化，还要加以充分恰当的引导和适度的调控，促进学生阅读水平的提高。 粗知大意并不是浅知。略读课文不像精读课文那么严格细致，只要求粗知文章大意。但这里所说的"粗知"是相对精读课文而言的，略读课文少了识字、学词学句等许多头绪，教总之，精读课文的教学，要紧扣教材的特点，精心设计教学方案，要注意发挥学生学习的积极性和主动性，引导学生进行探究、合作学习，要充分发挥教师的主导作用，要协调好个体学习与合作学习的关系，整体感悟和重点探究的关系，从而提高教学效率。而略读课文首先是在阅读内容上应"不求甚解"，在阅读技能培养上更强调运用。如叶老所说："就教学而言，精读是主体，略读只是补充；但是就效果而言，精读是准备，略读才是应用。"略的是教师的教，不能略的是学生的学。我们应该把"略读"提升到一个应有的高度，以求最大限度地扩张课堂，拓展课文，促进"学生语文素养的形成和发展"。 谈略读课文的教学方法 人教版教材中从三年级开始就有略读课文，三上的32篇课文中有8篇是略读，占总数的25%；四上的32篇课文中有14篇是略读课文，约占总数的44%；五六年级则占到了50%左右。从中，我们可以看出编辑的用心。但是，很多老师把略读课文上成精读课，也存在着略读课文怎么上的问题，什么可以略，什么不可以略？略读课文可以上成精读课文吗？还有，学生的阅读能力不高，一个重要的原因是：教学中，略读课文把握的不好。这些问题都困扰着我们，所以，略读课文的专项研讨变得非常重要。 叶圣陶先生曾经指出：“就教学而言，精读是主体，略读只是补充；但是就效果而言，精读是准备，略读才是应用。”“如果只注意于精读，而忽略了略读，功夫便只做得一半。”这些话十分精辟地阐述了精读与略读的关系，道出了略读课文教学的重要功能。《语文课程标准》也指出：“学会运用多种阅读方法。”“加强对阅读方法的指导，让学生逐步学会精读、略读和浏览。”一、明确目标 人教版课标实验教材对略读课的教学要求是：内容上，理解内容的要求要低于精读课文，一般是“粗知文章大意”，只要抓住重点、难点，帮助学生大体理解课文内容即可，词句的理解不作为重点;方法上，教师要更加放手，要让学生运用在精读课文学习中获得的知识与方法，自己把课文读懂，在实践中掌握读书方法，提高阅读能力。二、因“材”施教 教材虽然不是唯一的教学资源，但它是重要的教学资源。 1.根据课文结构因“材”施教。每一篇略读课文前都有阅读提示。阅读提示从激发阅读兴趣入手，一般提出一两个问题，侧重引导学生理解主要内容、揣摩教学思路，或引导学生从所阅读的文章扩展开去。对大多数略读课文来讲，教师根据阅读提示组织教学活动，皆可做到“提纲挈领”；学生根据阅读提示自主阅读学习。 2.根据课文特点因“材”施教每一篇略读课文，皆有它自己的特点。我们应根据课文自身的特点展开教学，既遵循文本正确的价值取向，又注意提高学生的学习兴趣。三、把握好教学的尺度 在略读课文的教学中，要把握好尺度，既不搞过细剖析，也不作随意教学，要确立略读课文教学的明确地位，发挥其应有的作用。教学中尤其要注意以下几点： 1.让学生自主学习。 略读课文要突出“略”字，略的是教师的精讲细说，强调“以学生自读为主”，让学生唱“主角”。教师不要包办代替，不要强行“灌输”，充分信任学生。教师甚至可以组织学生围绕某个重点、难点问题展开讨论、辩论，充分展示学生的独特思维和阅读体验，从而使学生既获得了知识，又锻炼了能力。 2.重“导”轻“教” 自读课文的教学强调“以学生的自读为主”，并不意味着忽视教师的“主导”作用。相反，略读课要真正达到“略”，教师就必须更加注重引导和调控。为了使教学中充分体现出“导”，教师应策划好课堂教学方法，及时、准确地评价学生的表现，激发学生的兴趣和自信心。 同时，略读课强调的是让学生学会阅读，教师要教学生如何去快速阅读，如何把握文章内容，而不是教学生具体的字、词、句等实实在在的考试知识。重“导”轻“教”，不是让教师退居幕后，而是要求教师发挥主导作用，引领学生深入、有效地开展有层次的阅读。 3.略读教学不等于略读，粗知大意并不是浅知。 略读教学是一种教学活动，它承担着更为广泛的任务，不仅是让学生在略读实践中获取信息，更重要的是要让学生在不断的实践中学习、学会略读方法，从而培养学生的阅读能力。教学中要授之以法，“教是为了不教。” 浅谈生物信息学在生物医药方面的应用 生物信息学（Bioinformatics）是在生命科学的研究中，以计算机为工具对生物信息进行储存、检索和分析的科学。它是当今生命科学和自然科学的重大前沿领域之一，同时也将是21世纪自然科学的核心领域之一。其研究重点主要体现在基因组学（Genomics）和蛋白质组学（Proteomics）两方面，具体说就是从核酸和蛋白质序列出发，分析序列中表达的结构功能的生物信息。 具体而言，生物信息学作为一门新的学科领域，它是把基因组DNA序列信息分析作为源头，在获得蛋白质编码区的信息后进行蛋白质空间结构模拟和预测，然后依据特定蛋白质的功能进行必要的药物设计。基因组信息学，蛋白质空间结构模拟以及药物设计构成了生物信息学的3个重要组成部分。是结合了计算机科学、数学和生物学的一门多学科交叉的学科。它依赖计算机科学、工程和应用数学的基础，依赖实验和衍生数据的大量储存。他将各种各样的生物信息如基因的DNA序列、染色体定位、基因产物的结构和功能及各种生物种间的进化关系等进行搜集、分类和分析，并实现全生命科学界的信息资源共享。 从生物信息学研究的具体内容上看，生物信息学可以用于序列分类、相似性搜索、DNA序列编码区识别、分子结构与功能预测、进化过程的构建等方面的计算工具已成为变态反应研究工作的重要组成部分。针对核酸序列的分析就是在核酸序列中寻找过敏原基因，找出基因的位置和功能位点的位置，以及标记已知的序列模式等过程。针对蛋白质序列的分析，可以预测出蛋白质的许多物理特性，包括等电点分子量、酶切特性、疏水性、电荷分布等以及蛋白质二级结构预测，三维结构预测等。 基因芯片是基因表达谱数据的重要来源。目前生物信息学在基因芯片中的应用主要体现在三个方面。 1、确定芯片检测目标。利用生物信息学方法，查询生物分子信息数据库，取得相应的序列数据，通过序列比对，找出特征序列，作为芯片设计的参照序列。 2、芯片设计。主要包括两个方面，即探针的设计和探针在芯片上的布局，必须根据具体的芯片功能、芯片制备技术采用不同的设计方法。 3、实验数据管理与分析。对基因芯片杂交图像处理，给出实验结果，并运用生物信息学方法对实验进行可靠性分析，得到基因序列变异结果或基因表达分析结果。尽可能将实验结果及分析结果存放在数据库中，将基因芯片数据与公共数据库进行链接，利用数据挖掘方法，揭示各种数据之间的关系。 大规模测序是基因组研究的最基本任务，它的每一个环节都与信息分析紧密相关。目前，从测序仪的光密度采样与分析、碱基读出、载体标识与去除、拼接与组装、填补序列间隙，到重复序列标识、读框预测和基因标注的每一步都是紧 四 川 大 学 网 络 教 育 学 院 模 拟 试 题( A ) 《管理运筹学》 一、 单选题（每题２分，共20分。） 1．目标函数取极小（minZ ）的线性规划问题可以转化为目标函数取极大的线性规 划问题求解，原问题的目标函数值等于（ C ）。 A. maxZ B. max(-Z) C. –max(-Z) D.-maxZ 2. 下列说法中正确的是（ B ）。 Ａ．基本解一定是可行解 Ｂ．基本可行解的每个分量一定非负 Ｃ．若B 是基，则B 一定是可逆Ｄ．非基变量的系数列向量一定是线性相关的 3．在线性规划模型中，没有非负约束的变量称为 （ D ） 多余变量 B ．松弛变量 C ．人工变量 D ．自由变量 4. 当满足最优解，且检验数为零的变量的个数大于基变量的个数时，可求得（ A ）。 Ａ．多重解 Ｂ．无解 Ｃ．正则解 Ｄ．退化解 5．对偶单纯型法与标准单纯型法的主要区别是每次迭代的基变量都满足最优检验但不完全满足 （ D ）。 A ．等式约束 B ．“≤”型约束 C ．“≥”约束 D ．非负约束 6. 原问题的第ｉ个约束方程是“＝”型，则对偶问题的变量i y 是（ B ）。 Ａ．多余变量 Ｂ．自由变量 Ｃ．松弛变量 Ｄ．非负变量 7.在运输方案中出现退化现象，是指数字格的数目( C )。 A.等于m+n B.大于m+n-1 C.小于m+n-1 D.等于m+n-1 8. 树Ｔ的任意两个顶点间恰好有一条（ B ）。 Ａ．边 Ｂ．初等链 Ｃ．欧拉圈 Ｄ．回路 9．若G 中不存在流f 增流链，则f 为G 的 （ B ）。 A ．最小流 B ．最大流 C ．最小费用流 D ．无法确定 10.对偶单纯型法与标准单纯型法的主要区别是每次迭代的基变量都满足最优检验但不完全满足（ D ） Ａ．等式约束 Ｂ．“≤”型约束 Ｃ．“≥”型约束 Ｄ．非负约束 二、多项选择题（每小题4分，共20分） 1．化一般规划模型为标准型时，可能引入的变量有 （ ） A ．松弛变量 B ．剩余变量 C ．非负变量 D ．非正变量 E ．自由变量 2．图解法求解线性规划问题的主要过程有 （ ） A ．画出可行域 B ．求出顶点坐标 C ．求最优目标值 D ．选基本解 E ．选最优解 3．表上作业法中确定换出变量的过程有 （ ） A ．判断检验数是否都非负 B ．选最大检验数 C ．确定换出变量 D ．选最小检验数 E ．确定换入变量 4．求解约束条件为“≥”型的线性规划、构造基本矩阵时，可用的变量有 （ ） A ．人工变量 B ．松弛变量 C. 负变量 D ．剩余变量 E ．稳态 变量 5．线性规划问题的主要特征有 （ ） A ．目标是线性的 B ．约束是线性的 C ．求目标最大值 D ．求目标最小值 E ．非线性 三、 计算题（共60分） 1. 下列线性规划问题化为标准型。(10分) 精读课文和略读课文的区别与教学重点 在小学语文教材中，精读课文的编排占据主体地位。就其涵盖的内容而言，包含了字词、句、段、篇等语文基本功训练项目，范围十分广泛，知识含量极高，是进行语文基本功训练的主要凭借。由此可见，精读课文的教学在小学语文教学中处于重中之重的地位，它是完成小学语文教学任务的重要途径。但是，就篇幅而言，略读课文在小学中段所占比例较小，但到了高段则是五五分成。因此略读课文的教学同样不可忽视。今天，我就以小学三年级的第一单元的教学为切入点来谈谈精读课文和略读课文教学的区别。 一、从教学目标上谈。 本单元文前三篇课文为精读课文，后一篇是略读课文，四篇课文从多角度地展现了大自然的美。我们可以从这些绿柳飞燕、鸟语花香、潺潺流水中，感受春天的来临，倾听大自然的声音，欣赏大自然的美景，从而激发热爱大自然的感情。我认真阅读了四篇课文，分析了教材说明，确立了： 精读课文的教学目标是： 1、加强词、句的训练。 2、重视朗读的指导和训练。朗读是最经常、重要的阅读训练，也是理解课文内容，也是体会课文思想感情的主要方法。 3、了解作者观察的方法，积累文中的优美句段。 阅读课文的教学目标： 1、用多种阅读方法获取重要信息，进行略读能力的训练，粗知文章的大意。略读课上，要更注重略读方法的运用，如扫读、跳读等，训练学生快速默读的方法，提高阅读的速度。 2、通过自主阅读获取信息，培养独立阅读能力，同时体现教师的引导作用。 3、加强课内外联系，沟通课内外阅读，适度拓展和延伸。略读课文的教学应该成为联系课内外阅读的桥梁，让学生在教师的指导下，尝试运用精读课．习得的方法独立阅读，实现向课外完全独立阅读的过渡，适当向课外拓展、延伸。 二、从教学过程上谈。 就精读课文《荷花》而言，可以这样教学： 1、精读。特级教师林润生是这样说的："精读课文的教学，应让学生拥有完整的阅读过程，经历精读与略读，朗读与默读，自悟与交流，思考与想象，课文阅读与资料收集利用，感知内容与领悟感情，理解与积累运用语言的过程。"他是这样设计《荷花》这一课的：课始让学生比赛谁对荷花了解的多少，课中通过五读课文（一读整体感知，二读领悟感情，三读再悟情感，四读升化情感，五读积累语句），让学生抓住重点词句朗读、揣摩、体会情感，联系上下文和联系有关资料深入领会重点语句的含义，深刻感受课文情感--大自然的优美。课末让学 四川大学网络教育学院模拟试题( A ) 《管理运筹学》 一、单选题（每题２分，共20分。） 1．目标函数取极小（minZ）的线性规划问题可以转化为目标函数取极大的线性规 划问题求解，原问题的目标函数值等于（C）。 A. maxZ B. max(-Z) C. –max(-Z) D.-maxZ 2.下列说法中正确的是（B）。 Ａ．基本解一定是可行解Ｂ．基本可行解的每个分量一定非负 Ｃ．若B是基，则B一定是可逆Ｄ．非基变量的系数列向量一定是线性相关的3．在线性规划模型中，没有非负约束的变量称为（ D ） 多余变量B．松弛变量C．人工变量D．自由变量 4. 当满足最优解，且检验数为零的变量的个数大于基变量的个数时，可求得 （ A ）。 Ａ．多重解Ｂ．无解Ｃ．正则解Ｄ．退化解5．对偶单纯型法与标准单纯型法的主要区别是每次迭代的基变量都满足最优检验 但不完全满足（ D ）。 A．等式约束 B．“≤”型约束 C．“≥”约束 D．非负约束 y是（ B ）。 6. 原问题的第ｉ个约束方程是“＝”型，则对偶问题的变量i Ａ．多余变量Ｂ．自由变量Ｃ．松弛变量Ｄ．非负变量 7.在运输方案中出现退化现象，是指数字格的数目( C )。 A.等于m+n B.大于m+n-1 C.小于m+n-1 D.等于m+n-1 8.树Ｔ的任意两个顶点间恰好有一条（B）。 Ａ．边Ｂ．初等链Ｃ．欧拉圈Ｄ．回路9．若G中不存在流f增流链，则f为G的（ B ）。 A．最小流 B．最大流 C．最小费用流 D．无法确定 10.对偶单纯型法与标准单纯型法的主要区别是每次迭代的基变量都满足最优检验 但不完全满足（ D ） Ａ．等式约束Ｂ．“≤”型约束Ｃ．“≥”型约束Ｄ．非负约束二、多项选择题（每小题4分，共20分） 1．化一般规划模型为标准型时，可能引入的变量有（） A．松弛变量 B．剩余变量 C．非负变量 D．非正变量 E．自由变量 2．图解法求解线性规划问题的主要过程有（） A．画出可行域 B．求出顶点坐标 C．求最优目标值 D．选基本解 E．选最优解 3．表上作业法中确定换出变量的过程有（） A．判断检验数是否都非负 B．选最大检验数 C．确定换出变量 D．选最小检验数 E．确定换入变量 4．求解约束条件为“≥”型的线性规划、构造基本矩阵时，可用的变量有（）A．人工变量 B．松弛变量 C. 负变量 D．剩余变量 E．稳态变量 5．线性规划问题的主要特征有（） A．目标是线性的 B．约束是线性的 C．求目标最大值 D．求目标最小值 E．非线性 三、计算题（共60分） 1. 下列线性规划问题化为标准型。(10分) 1 / 17 小学语文新课程标准教材 语文教案( 2019 — 2020学年度第二学期 ) 学校： 年级： 任课教师： 语文教案 / 小学语文 / 小学五年级语文教案 编订：XX文讯教育机构 精读与略读课文 教材简介:本教材主要用途为通过学习语文的内容，培养学生的阅读能力、交流能力，学习语文是为了更好的学习和工作，为了满足人类发展和实现自我价值的需要，本教学设计资料适用于小学五年级语文科目, 学习后学生能得到全面的发展和提高。本内容是按照教材的内容进行的编写，可以放心修改调整或直接进行教学使用。 一个人要积累知识，就必须读书。对重要的文章和书籍，要认真读反复读，要逐字逐句地深入钻研，对重要的语句和章节所表达的思想内容还要做到透彻理解。这就是精读。我们所读的课文，都是经过精心编选的，对我们学习语文很有帮助，因此须要精读。 然而，人的精力是有限的，而书籍却浩如烟海，谁也不可能读完所有的书，更不可能对每本书每篇文章都去精读。为了获取更多的知识，更多的信息，又必须读大量的书，因此我们还要学会略读。 略读就大略的读。略读可以增加阅读量。通过略读，可以用较少的时间浏览大量的书刊，从而扩大自己的知识面，获得对有关读物的总的认识，便于以后需要的时候去查找。略读又是精读的基础。通过泛读，可以在很短的时间内知道一篇文章或一本书的基本内容，从而确定它是否须要精读，或哪些地方需要精读。 怎样进行泛读呢？ 如果面对一大堆读物，我们可以将每本书的内容提要、前言、章节、目录等很快地从头 到尾看一遍，这样就能大致了解每本书的梗概、特点和应用范围。对于一本读物来说，可以几段几段地粗读，也可几页几页地翻阅，而不要探究某个字、词或句子的意思。俗话说的“一目十行”就是指这种走马观花式的快速阅读方法。 为了提高读书速度，略读通常采用默读的方式。 当然，泛读也并不是读完了事，每次泛读完毕，也应想一下所读的东西，最好用简炼的语言概括一个读物内容。 只要我们善于把精读和略读结合起来，就能取得最佳的读书效果。 XX文讯教育机构 WenXun Educational Institution 例谈小学略读课文教学策略 长久以来，略读课文在语文教学中处于比较尴尬的境地：一方面，它是课文中的组成部分，不教不行；另一方面，正因为它是“略读”课文，因此，在教学中往往被大多教师所忽视，在实际教学中，教师往往轻描淡写，一读而过，学生对课文也没留下多大印象。长此以往，限制了学生的阅读能力发展。对此，我们必须重新审视略读课文教学，采取有效的教学策略，使略读课文教学达到它应有的目标。以下是我在略读课文教学过程中的一些想法和采取的策略。 策略一：紧抓提示，凸显重点 我们知道略读课文一般只有一课时，在这短短的一课时中，教师快速有效地抓准略读课文的教学重点，使略读课文和精读课文形成一个系统，形成教学合力，更好地发挥训练阅读，迁移能力和陶冶情操的功能。如何抓？从哪里抓？每篇略读课文前的导语为我们提供了很好的借鉴。人教版实验教材在每一课略读课文前都有一段承上启下的提示语，如五年级下册册《丝绸之路》的提示语是这样写的： 祖国的西部有美丽的自然风光，还有灿烂的历史文化。著名的丝绸之路，就是其中光辉的一页。读读下面这篇课文，想想课文讲了哪些内容，从中体会到了什么。有条件的还可以搜集有关丝绸之路的故事和同学交流。 这段提示语优美生动，既激发了学生的阅读兴趣，延续了本组课文的主题，又明确提出了这篇课文的阅读要求。我们可以把这段提示语概括为两个方面： 一、读课文，了解课文讲了哪些内容，从中体会到了什么。 这是要求我们在教学初始引导学生整体把握课文内容，了解文中介绍的主要内容和历史意义，激发学生热爱祖国西部的情感。 二、搜集有关丝绸这路的故事和同学交流。 这点则是要求我们在学完课文的同时，能够引导学生课后去搜集资料，培养学生搜集信息的能力，同时也丰富学生的课外知识，增进与同学间的合作。 像这样，对提示语进行二次解读，既有助于我们教师找准重点，又能及时有效地展开略读课文的教学。 策略二：注重预习，培养自主 苏霍姆林斯基曾经说过这样一段精辟的论语：“在人的心灵深处，都有一种根深蒂固的需要，这就是希望自己是一个发现者、研究者、探索者。在儿童的精神世界里，这种需要特别强烈。”在教学中，教师应根据学生的这一心理特点，激发学生的学习兴趣，培养自主学习的能力。 《运筹学》试题样卷（一） 一、判断题（共计10分，每小题1分，对的打√，错的打X ） 1. 无孤立点的图一定是连通图。 2. 对于线性规划的原问题和其对偶问题，若其中一个有最优解， 另一个也一定有最优解。 3. 如果一个线性规划问题有可行解，那么它必有最优解。 4．对偶问题的对偶问题一定是原问题。 5．用单纯形法求解标准形式（求最小值）的线性规划问题时，与0 >j σ对应的变量都可以被选作换 入变量。 6．若线性规划的原问题有无穷多个最优解时，其对偶问题也有无穷 多个最优解。 7. 度为0的点称为悬挂点。 8. 表上作业法实质上就是求解运输问题的单纯形法。 9. 一个图G 是树的充分必要条件是边数最少的无孤立点的图。 二、建立下面问题的线性规划模型（8分） 某农场有100公顷土地及15000元资金可用于发展生产。农场劳动力情况为秋冬季3500人日；春夏季4000人日。如劳动力本身用不了时可外出打工，春秋季收入为25元 / 人日，秋冬季收入为20元 / 人日。该农场种植三种作物：大豆、玉米、小麦，并饲养奶牛和鸡。种作物时不需要专门投资，而饲养每头奶牛需投资800元，每只鸡投资3元。养奶牛时每头需拨出1.5公顷土地种饲料，并占用人工秋冬季为100人日，春夏季为50人日，年净收入900 元 / 每头奶牛。养鸡时不占用土地，需人工为每只鸡秋冬季0.6人日，春夏季为0.3人日，年净收入2元 / 每只鸡。农场现有鸡舍允许最多养1500 三、已知下表为求解某目标函数为极大化线性规划问题的最终单纯形表，表中54 ,x x 为松弛变量，问 (1)写出原线性规划问题；（4分） (2)写出原问题的对偶问题；（3分） (3)直接由上表写出对偶问题的最优解。（1分） 四、用单纯形法解下列线性规划问题（16分） s. t. 3 x1 + x2 + x3?60 x 1- x 2 +2 x 3?10 x 1+x 2-x 3?20 x 1,x 2 ,x 3?0 五、求解下面运输问题。（18分） 某公司从三个产地A1、A2、A3将物品运往四个销地B1、B2、B3、B4，各产地的产量、各销地的销量和各产地运往各销地每件物品的运费如表所示： 问：应如何调运，可使得总运输费最小? 六、灵敏度分析（共8分） 线性规划max z = 10x1 + 6x2 + 4x3 s.t. x1 + x2 + x3 ?100 10x1 +4 x2 + 5 x3 ?600 2x1 +2 x2 + 6 x3 ?300 x1 , x2 , x3 ?0 的最优单纯形表如下： (1)C1在何范围内变化，最优计划不变？(4分) (2)b1在什么范围内变化，最优基不变？(4分) 七、试建立一个动态规划模型。（共8分） 略读课与精读课的区别 关于略读课，始于2008年冬天的大讨论已历时近两年，但是，无论是各级名师的观摩课、示范课，还是一般教师的公开课，或是刊物上发表的教学设计、教学实录，只要公开展示，就几乎“涛声依旧”——上成“原版”或“缩微版”的精读课。为何？首先，对于略读课的性质、定位存在分歧或模糊认识；其次，即便认识一致将其落实于实践也需要一个过程；再次，还有另一个重要原因不容忽视，那就是惯性的作用——轻车熟路地套用精读课的教学模式。眼下，继续澄清对于略读课的模糊认识有其必要性，而为其构建易于为广大一线教师所掌握的教学模式更必须摆上日程。构建略读课的教学模式固然应以正确认识其性质、定位为先决条件和起点，但克服惯性作用也至关重要，否则无法摆脱业已驾轻就熟的教学模式的影响，也就不会有真正意义上的略读课。诚然，“教无定法”，略读课不能囿于一式，但不管何种模式，都应在以下诸方面别于精读课： 顾名思义，人们往往根据“精读”和“略读”的涵义去界定“精读课”和“略读课”的性质、定位，这虽符合一般逻辑，在这里却是实实在在的望文生义。从上个世纪八十年代起，语文教材就有一类课文和二类课文，在中学分别称为“讲读课文”和“自读课文”，在小学分别称为“讲读课文”和“阅读课文”。进入2000年以后，部分课标本实验教材分别称为“教读课文”和“自读课文”或者“主体课文”和“拓展阅读课文”，而多数则分别改称为“精读课文”和“略读课文”。二类课文性质为何？1986年版《小学语文教学大纲》指出：“对阅读课文，教师要着重指导学生运用从讲读课中学到的学习方法进行阅读。”明确其半独立阅读的性质。2000年小学语文大修订版教材出版，由人教社小语室编著的《教师用书》指出：“略读课文具有较明显的独立阅读的性质。安排略读课文，主要是引导学生把从精读课文中学到的语文基本功，用于阅读实践，逐步培养独立阅读能力。”命名虽异，半独立阅读的性质不变。一、二课文是相对的关系，二类课文的“半独立”，意味着一类课文的“不独立”。“不独立”和“半独立”显示教师和学生在教学中的地位和作用有异，表明教师指导和学生参与的程度有别，而这正决定了二者性质和定位的不同。在两类课文所有的名称中，“教读课文”和“自读课文”彰显了师生在其中的不同地位和作用，比较准确地揭示了二者的性质和定位。而“精读”和“略读”是两种粗细不同的阅读方法，以此命名课 运筹学期末试卷（B卷） 系别：工商管理学院专业：考试日期：年月日姓名：学号：成绩： 1．[10分] 匹克公司要安排4个工人去做4项不同的工作，每个工人完成各项工作所消耗的时间（单位：分钟）如下表所示： 要求：（1）建立线性规划模型(只建模型,不求解) （2）写出基于Lindo软件的源程序。 2.[15分]某公司下属甲、乙两个厂，有A原料360斤，B原料640斤。甲厂用A、B两种原料生产x1,x2两种产品，乙厂也用A、B两种原料生产x3,x4两种产品。每种单位产品所消耗各种原料的数量及产值、分配等如下 （1） 建立规划模型获取各厂最优生产计划。 （2） 试用图解法 求解最优结果。 3．[10分] 考虑下面的线性规划问题: 目标函数：Min Z=16x 1+16x 2 +17x 3 约束条件： 利用教材附带软件求解如下: **********************最优解如下************************* 目标函数最优值为 : 148.916 变量 最优解 相差值 ------- -------- -------- x1 7.297 0 x2 0 .703 x3 1.892 0 约束 松弛/剩余变量 对偶价格 ------- ------------- -------- 13123123123300.56153420,,0 x x x x x x x x x x x +≤-+≥+-≥≥ 1 20.811 0 2 0 -3.622 3 0 -4.73 目标函数系数范围: 变量下限当前值上限 ------- -------- -------- -------- x1 1.417 16 16.565 x2 15.297 16 无上限 x3 14.4 17 192 常数项数范围: 约束下限当前值上限 ------- -------- -------- -------- 1 9.189 30 无上限 2 3.33 3 15 111.25 3 -2.5 20 90 试回答下列问题： （1）第二个约束方程的对偶价格是一个负数(为-3.622)，它的含义是什么? （2）x2有相差值为0.703,它的含义是什么? （3）请对右端常数项范围的上、下限给予具体解释，应如何应用这些数 对略读课文教学的实践与思考长久以来，由于受应试教育的影响（略读课文不列入考试范围），许多老师只重视精读课文的教学，对略读课文的教学研究甚少，不能很好地把握它的特点，在实际的教学中，往往出现两种倾向，要么与精读课文的教学没有什么区别，字词句段、篇章结构，面面俱到、精雕细琢；要么就轻描淡写、囫囵吞枣，一读带过，学生根本没有留下什么印象。那么，略读课文教学的目标究竟又是什么？教学中又如何体现它的教学特点呢？我们通过对人教版三上《一幅名扬中外的画》一课的教学实践，有了一些思考，下面谈谈对略度课文教学的认识。 一、领会略读课文的编排意图 叶圣陶曾经说过：“就教学而言，精读是主体，略读只是补充；但是就效果而言，精读是准备，略读才是应用。”“如果只注意于精读，而忽略了略读，功夫便只做得一半”。“精读文章，只能把它认作例子与出发点，既熟习了例子，只定了出发点，就得推广开来，阅读略读书籍。”叶老已经十分精辟地阐述了精读与略读的关系，精读是略读的基础，略读是精读的补充，它们都是阅读的最基本的方法，有着各自不同的作用，略读与精读一样重要。《语文课程标准》也指出：“学会运用多种阅读方法。”“加强对阅读方法的指导，让学生逐步学会精读、略读和浏览。”这里同样提出了略读的要求。其实，在信息时 代，略读更能迅速、便捷地获取大量信息，在日常生活中，略读比精读应用更为广泛。 以上着力强调了略读的重要性，有利于我们正确领会略读课文教学的编写意图。人教版新课程实验教材从二年级下册开始安排略读课文，并提出了对这类课文教学的要求：一是从内容上说，理解课文的要求要低于精读课文的教学，一般是“粗知课文大意”，只要抓住重点、难点帮助学生理解即可，词句的理解不作为训练的重点。二是从方法上说，教师要更加放手，主要靠学生自己把课文读懂，并在读中掌握读书方法，提高阅读能力。由此可见，编者安排略读课文的教学，目的之一是要我们训练学生略读的能力，不要求咬文嚼字，只要求“粗知课文大意”；目的之二是让我们培养学生独立阅读的能力，自己把课文读懂，在实践中掌握读书方法。二、明确略读课文教学的目标 基于以上对略读课文教学的认识，它的主要教学目标可以为： 1．用多种阅读方法获取重要信息，尤其要进行略读能力的训练，读懂文章的大意。略读课上，要更注重略读方法的运用，如扫读、跳读等，训练学生快速默读的方法，提高阅读的速度，当然也不排斥精读方法的融合运用。 2．通过自主阅读获取信息，培养独立阅读能力。略读课是学生独立阅读的实践机会，个体自读和合作交流是略读课文 《管理运筹学》期末考试试题 一、单项选择题（共5小题，每小题3分，共15分） 1.如果一个线性规划问题有n个变量，m个约束方程(m 3. 写出下面线性规划问题的对偶问题： 123123123123123min z 25, 258, 23 3,.. 4 26, ,,0. x x x x x x x x x s t x x x x x x =++-+≤??++=??-+≤??≥? 四、计算下列各题（每题20分，合计40分） 1. 用单纯形法求解下列线性规划的最优解： 012121212max 2..32250,0x x x s t x x x x x x =+??≤??≤??+≤??≥≥? 2.用割平面法求解整数规划问题。 12 121212 max 7936735,0,z x x x x x x x x =+-+≤??+≤??≥?且为整数 如何处理精读课文和略读课文 略读课作为一种课程形态出现在教材中，主要目的就是为了培养学生的略读能力。因此，略读课的教学应该重视略读方法的训练。著名特级教师王燕骅老师说：“略读课文的教学应当是…教?略，而学要…丰?；应当是…教?略，而…学?不略。”叶圣陶先生也谈过，略读的略字，一半系就教师的指导而言，只须提纲挈领的指导；一半系就学生的功夫而言，还是要像精读那样仔细咬嚼……所以，略读课文教学的追求是──在“精”与“略”之间行走，重其所重，略其所略。整体把握求“略”，教师指导和学生所花的时间略；重点感悟求“精”，对于文章的重点、精彩之处还是要引导学生细细品读。 “精略相辅，略中有精，…教?略而…学?不略”，略读课文教学定位的确定，正为我们略读教学的探索指明了方向，我们认为可用如下方法进行教学。 （一）整体感知求“略” 对课文的内容、词句、结构、表达方法，在整体感知的阶段，力求放开，从略处理。只求尽量快地尽量多地略读浏览，搜集信息，多角度、多方位地粗知文章大意，对全文形成一个比较粗略的整体认识。 1.整体性阅读。整体性阅读是对于课文从整体上快速理解的阅读方法，速读的主要程序是了解文章题目、人物（景物）、时间、地点、主要事件（景物特点）、自己的感受体会等。 如请同学们根据阅读提示，采用自己喜欢的方式阅读课文，力求读通读顺；对长的、难读的句子可以多读几遍；同时，边读边想阅读提示中的问题，然后尝试利用查字典等方法解决问题，在自己感受最深和感觉还有问题的地方做上记号。 2.寻找式阅读。是寻找特定信息的一种阅读方法。学生可以根据阅读提示所提示的问题寻找信息，也可以根据自己在整体性阅读中所发现的问题寻找信息。 略读课文前面的“阅读提示”，除了用于激发学生的阅读兴趣外，还着重对课文的主旨和阅读方法、学习的重难点、语文实践的拓展点给予提纲挈领性的导向。因此要引导学生对照“阅读提示”进行自主读书——或朗读，或默读，或高声诵读，或轻声低吟，或就精彩片断反复品读，或对部分描述扫视而过，从而粗知大意。 3.浏览式阅读。只求了解大体内容而不顾细枝末节，不阅读全文就能掌握材料内容的一种阅读方法。浏览式阅读法的程序：①看目录或标题，选择自己感兴趣的内容。②浏览正文，有选择性地阅读文章中的某些部分。③整理收获。读后概括中心思想、文章要点，以形成总的印象。 读通课文，整体把握是略读教学很重要的一步，为下一步理解课文作铺垫。只有读准了，读通了，读顺了，才会有下一步的“我自己读懂了”。 （二）重点感悟求“精” 编者定为略读课文，定有其道理，我们把握时要注意这一点，与精读课文有所区别，但也要尽可能觅到“精教精学”的乐趣。在略读整体感知的基础上，对课文特别是“阅读提示”所要求的重点进行探究时要采用精读方法，通过小组学习，通过研读形式，进行比较集中深入的感悟，追求──“精略相辅，略中有精，教略而学不略”。 1.主线点拨法。从方法上说，略读课文就是让孩子把在精读课文中学习到的方法运用到自己的学习中，并在读中掌握读书方法，提高阅读能力。教师的作用就在于适当点拨和引导，使学生的个性在课堂中张扬，阅读水平得到提高。我们认为围绕主线，精妙点拨是有效的方法。 2.重点突破法。人教版教材的略读课文中有不少文章和精读课文一样，文质兼美，情趣突出，可以深入文字，在教学中寻找适当的突破口、“感悟点”进行语言文字训练，引导学生细细品读文章的重点、精彩之处。 3.语言积累法。文字很美的文章，则注重语言积累。“美文诵读”是理解、感悟课文的捷径。学生在诵读中，积累了景物描写的语言和手法；在品词品句中，自然而然地就把这些好词佳句记忆在脑中，丰富了自己的“语言仓库”。精读课文和略读课文的授课区别
学习略读课文方法
浅谈生物信息学在生物医药方面的应用
管理运筹学模拟试题及答案
精读课文与略读课文的区别
管理运筹学模拟试题附答案
五年级语文：精读与略读课文
例谈小学略读课文教学策略
《运筹学》期末考试试卷A答案
略读课与精读课的区别
管理运筹学期末试卷题目B卷
对略读课文教学的实践与思考
《管理运筹学》期末考试试题
如何处理精读课文和略读课文
相关主题
文本预览