Bioinformatics
- 2. Introduction to Bioinformatics Combined to solve complex Biological problems Algorithms and techniques of computer science being used to solve the problems faced by molecular biologists ‘Information technology applied to the management and analysis of biological data’ Storage and Analysis are two important functions – bioinformaticians build tools for each Bio IT market has observed significant growth in genomic era Biology Chemistry Statistics Computer science Bioinformatics
- 3. Fields of Bioinformatics The need for bioinformatics has arisen from the recent explosion of publicly available genomic information, such as resulting from the Human Genome Project. Gain a better understanding of gene analysis, taxonomy, & evolution. To work efficiently on the rational drug designs and reduce the time taken for development of drug manually. Unravel the wealth of Biological information hidden in mass of sequence, structure, literature and biological data Has environmental-clean up benefits In agriculture, it can be used to produce high productivity crops Gene Therapy Forensic Analysis Understanding biological pathways and networks in System Biology
- 4. Bioinformatics key areas Bioinformatics lecture March 5, 2002 organisation of knowledge (sequences, structures, functional data) e.g. homology searches
- 5. Applications of Bioinformatics Provides central, globally accessible databases that enable scientists to submit, search and analyze information and offers software for data studies, modelling and interpretation. Sequence Analysis:- The application of sequence analysis determines those genes which encode regulatory sequences or peptides by using the information of sequencing. These computational tools also detect the DNA mutations in an organism and identify those sequences which are related. Special software is used to see the overlapping of fragments and their assembly. Prediction of Protein Structure:- It is easy to determine the primary structure of proteins in the form of amino acids which are present on the DNA molecule but it is difficult to determine the secondary, tertiary or quaternary structures of proteins. Tools of bioinformatics can be used to determine the complex protein structures. Genome Annotation:- In genome annotation, genomes are marked to know the regulatory sequences and protein coding. It is a very important part of the human genome project as it determines the regulatory sequences
- 6. Comparative Genomics:- Comparative genomics is the branch of bioinformatics which determines the genomic structure and function relation between different biological species which enable the scientists to trace the processes of evolution that occur in genomes of different species. Pharmaceutical Research:- Tools of bioinformatics are also helpful in drug discovery, diagnosis and disease management. Complete sequencing of human genes has enabled the scientists to make medicines and drugs which can target more than 500 genes. Accurate prediction in screening.
- 7. S. No Unix Windows Linux 1. Open source Close source Open source 2. Very high security system Low security system High security system 3. Command-line GUI Hybrid 4. File system is arranged in hierarchical manner File system is arranged in parallel manner File system is arranged in hierarchical manner 5. Not user friendly User friendly User friendly 6. Single tasking Multi tasking Multi tasking
- 9. Biological databanks and databases Very fast growth of biological data Diversity of biological data: o Primary sequences o 3D structures o Functional data Database entry usually required for publication o Sequences o Structures Nucleic Acid Protein EMBL (Europe) PIR - Protein Information Resource GenBank (USA) MIPS DDBJ (Japan) SWISS-PROT University of Geneva, now with EBI TrEMBL A supplement to SWISS- PROT NRL-3D Major primary databases
- 10. Sequence Databases Three databanks exchange data on a daily basis Data can be submitted and accessed at either location Nucleotides db: GenBank - https://www.ncbi.nlm.nih.gov/ EMBL - https://www.ebi.ac.uk/ DDBJ - https://www.ddbj.nig.ac.jp/index-e.html Bibliographic db: PubMed , Medline Specialized db: RDP, IMGT, TRANSFAC, MitBase Genetic db: SGD – https://www.yeastgenome.org/ ACeDB, OMIM
- 11. Composite Databases Secondary Databases Swiss Prot PIR GenBank NRL-3D Store structure info or results of searches of the primary databases Composite Databases Primary Source PROSITE https://prosite.expasy.org/ SWISS-PROT PRINTS http://130.88.97.239/PRINTS/index.p hp OWL
- 12. Biological Database Carbohydrate Structure Database: CCSD – https://cordis.europa.eu/project/id/BIOT0184 https://www.genome.jp/dbget-bin/www_bfind?carbbank Glycome DB – http://www.glycome-db.org/ Metabolic or Enzyme Databases BRENDA – https://www.brenda-enzymes.org/ KEGG – https://www.genome.jp/kegg/ Structure Databases: CATH – https://www.cathdb.info/ SCOP- http://scop.mrc-lmb.cam.ac.uk/
- 13. SCOP Structural Classification of Proteins http://scop.mrc-lmb.cam.ac.uk/ SCOP database aims to provide a detailed and comprehensive description of structural and evolutionary relationship between all proteins Levels of hierarchy Family : Pairwise residue identities of aa 30% or greater Superfamily : Eventhough low seq identities, should have common evolutionary origin Eg: ATPase domain of HSP and HK Fold : Major structural similarity Class : all α , all β, α or β, α and β, Multidomain
- 14. CATH https://www.cathdb.info/ Class : 2º structure Architecture : Gross orientation of 2º structure, independent of connectivities Topology (fold family) : topological connection of super families S level : Sequence and structural identities
- 15. Basis of Sequence Alignment 1. Aligning sequences 2. To find the relatedness of the proteins or gene, if they have a common ancestor or not. 3. Mutation in the sequences, brings the changes or divergence in the sequences. 4. Can also reveal the part of the sequence which is crucial for the functioning of gene or protein. Similarity indicates conserved function Human and mouse genes are more than 80% similar Comparing sequences helps us understand function
- 16. Sequence Alignment After obtaining nuc/aa sequences, first thing is to compare with the known sequences. Comparison is done at the level of constituents. Then finding of conserved residues to predict the nature and function of the protein. This process of mapping is called Sequence Alignment 1. Local alignment – Smith & Waterman Algorithm 2. Global alignment – Needleman & Wunch Algorithm Gapped Alignment Ungapped Alignment Terms to Know - Homolog, Ortholog, Paralog, Xenolog, Similar and Identical Alignment scoring and substitution matrices Dot plots Dynamic programming algorithm Heuristic methods (In order to reduce time) FASTA BLAST Pairwise sequence alignment Multiple sequence alignment
- 17. Scoring a sequence alignment Match score: Mismatch score: Gap penalty: + 1 + 0 –1 Matches: 18 × (+1) Mismatches: 2 × 0 Gaps: 7 × (– 1) Score = +11 ACGTCTGAT-------ATAGTCTATCT ACGTCTGATACGCCGTATAGTCTATCT AC-T-TGA--CG-CGT-TA-TCTATCT We can achieve this by penalizing more for a new gap, than for extending an existing gap Maximum no of matches gives high similarity – Optimum Alignment ACGTCTGATACGCCGTATAGTCTATCT ||||| ||| || |||||||| ----CTGATTCGC---ATCGTCTATCT
- 18. Scores: positive for identical or similar negative for different negative for insertion in one of the two sequences Substitution matrices – weights replacement of one residue by another assumption of evolution by point mutations amino acid replacement (by base replacement) amino acid insertion amino acid deletion Significance of alignment Depends critically on gap penalty Need to adjust to given sequence
- 19. Derivation of substitution matrices PAM matrices First substitution matrix; Developed by Dayhoff (1978) based on Point Accepted Mutation (PAM) model of evolution 1PAM (without sub) is a unit of evolutionary divergence in which 1% of the aa have been changed Derived from alignment of very similar sequences PAM1 = mutation events that change 1%of AA PAM2, PAM3, ... extrapolated by matrix multiplication e.g.: PAM2 = PAM1*PAM1; PAM3 = PAM2 * PAM1 etc Lower distance PAM matrix for closely related proteins eg., PAM30 Higher distance PAM matrix for highly diverged sequences eg., PAM250 Problems with PAM matrices: Incorrect modelling of long time substitutions, since conservative mutations dominated by single nucleotide change e.g.: L <–> I, L <–> V, Y <–> F long time: any Amino Acid change
- 20. positive and negative values identity score depends on residue positive and negative values identity score depends on residue
- 21. BLOSUM matrices BLOCKSAmino acid Substitution Matrices Similar as PAM; however the data were derived from local alignments for distantly related proteins deposited in BLOCKS db Unlike PAM there is no evolutionary basis BLOSUM series (BLOSUM50, BLOSUM62, ...) BLOCKS database: ungapped multiple alignments of protein families at a given identity E.g., BLOSUM 30 better for gapped alignments – for comparing highly diverged seq BLOSUM 90 better for ungapped alignments – for very close seq BLOSUM 62 was derived from a set of sequences which are 62% or less similar
- 22. DOT Plot Simple comparison without alignment 2D graphical representation method primarily used for finding regions of local matches between two sequences DOTTER, PALIGN, DOTLET (https://dotlet.vital-it.ch/) Distinguish by alignment score Similarities increase score (positive) Mismatches decrease score (Negative) Gaps decrease score Number of possible dots = (probability of pair) x (length of seq A) x (length of seq B) Disadv – No direct seq homology & Statistically weak
- 23. Dynamic programming algorithm To build up optimal alignment which maximizes the similarity we need some scoring methods The dynamic programming relies on a principle of optimality. PROCEDURE Construct a two-dimensional matrix whose axes are the two sequences to be compared. The scores are calculated one row at a time. This starts with the first row of one sequence, which is used to scan through the entire length of the other sequence, followed by scanning of the second row. The scanning of the second row takes into account the scores already obtained in the first round. The best score is put into the bottom right corner of an intermediate matrix. This process is iterated until values for all the cells are filled.
- 24. Depicting the results: Back tracing The best matching path is the one that has the maximum total score. If two or more paths reach the same highest score, one is chosen arbitrarily to represent the best alignment. The path can also move horizontally or vertically at a certain point, which corresponds to introduction of a gap or an insertion or deletion for one of the two sequences.
- 25. BLAST Basic Local Alignment search tool https://blast.ncbi.nlm.nih.gov/Blast.cgi Multi-step approach to find high-scoring local alignments between two sequences List words of fixed length (3AA) (11nuc) expected to give score larger than threshold (seed alignment) For every word, search database and extend ungapped alignment in both directions upto a certain length to get HSPs New versions of BLAST allow gaps Blastn: Blastp: tBlastn: Blastx: tBlastx: nucleotide sequences protein sequences protein query - translated database nucleotide query - protein database nucleotide query - translated database
- 27. Interpretation Rapid and easier to find homolog by scanning huge db Search against specialized db Blast program employ SEG program to filter low complexity regions before executing db search Quality of the alignment is represented by score (to identify hits) Significance of the alignment is represented as e-value (Expected value) E-value decreases exponentially as the score increases The E-value provides information about the likelihood that a given sequence match is purely by chance. The lower the E- value, the less likely the database and therefore more significant the match is. If E is between 0.01 and 10, the match is considered not significant.
- 28. FASTA More sensitive than BLAST Table to locate all identically matching words of length Ktup between two sequences Blast – Hit extension step Fasta – Exact word match As the high value of Ktup increases the search becomes slow FASTA also uses E-values and bit scores. The FASTA output provides one more statistical parameter, the Z-score. If Z is in the range of 5 to 15, the sequence pair can be described as highly probable homologs. If Z < 5, their relationships is described as less certain
- 29. Phylogenetics Phylogenetics is the study of evolutionary relatedness among various groups of organisms (e.g., species, populations). Methods of Phylogenetic Analysis: Monophyletic group – all taxa share by one common ancestor Paraphyletic groping – share common ancestor but not all Errors in alignment mislead tree Phenetic NJ, UPGMA Cladistic MP ML
- 30. A phylogenetic tree is a tree showing the evolutionary interrelationships among various species or other entities that are believed to have a common ancestor. A phylogenetic tree is a form of a cladogram. In a phylogenetic tree, each node with descendants represents the most recent common ancestor of the descendants, and edge lengths correspond to time estimates. Each node in a phylogenetic tree is called a taxonomic unit. Internal nodes are generally referred to as Hypothetical Taxonomic Units (HTUs) as they cannot be directly observed Distances – no of changes Parts of a phylogenetic tree Node Root Outgroup Ingroup Branch
- 31. Phenetic Method of analysis: Also known as numerical taxonomy Involves various measures of overall similarity for ranking species All the data are first converted to a numerical value without any character (weighing). Then no of similarities / differences is calculated. Then clustering or grouping close together Lack of evolutionary significance in phenetics Cladistic method of analysis: Alternative approach Diagramming relationship between taxa Basic assumption – members of the group share a common evolutionary history Typically based on morphological data
- 32. Distance and Character A tree can be based on 1. quantitative measures like the distance or similarity between species, or 2. based on qualitative aspects like common characters. Molecular clock assumption – substitution in nu / aa are being compared at constant rate
- 33. Maximum Parsimony: Finds the optimum tree by minimizing the number of evolutionary changes No assumptions on the evolutionary pattern MSA then scoring Rather time consuming works well if seq have strong similarity May oversimplify evolution May produce several equally good trees PAUP, MacClade Maximum Likelihood: The best tree is found based on assumptions on evolution model Nucleotide models more advanced at the moment than aminoacid models Programs require lot of capacity from the system
- 34. Neighbour Joining: The sequences that should be joined are chosen to give the best least-squares estimates of the branch length that most closely reflect the actual distances between the sequences NJ method begins by creating a star topology in which no neighbours are connected Then tree is modified by joining pair of sequences. Pair to be joined is chosen by calculating the sum of branch length Distance table No molecular clock assumed UPGMA Unweighted Pair Group method with Arithmetic Mean Works by clustering, starting with more similar towards distant Dot representation Molecular clock assumed
- 35. PHYLIP (Phylogeny Inference Package) Available free in Windows/MacOS/Linux systems Parsimony, distance matrix and likelihood methods (bootstrapping and consensus trees) Data can be molecular sequences, gene frequencies, restriction sites and fragments, distance matrices and discrete characters