BITS - Comparative genomics: gene family analysis
3 likes1,980 views
This document discusses comparative genomics and gene family analysis. It describes clustering proteins into families to study evolution, orthology, paralogy, gene duplication and loss. Gene family analysis allows detection of errors in gene structure annotation by comparing sequences. Phylogenetic trees classify gene families and reveal functional divergence. Resources for orthology analysis include Ensembl, OrthoMCLDB and YGOB. The goal of hands-on analysis is to characterize the talin 2 gene family.
![Examples of family-specific protein
motifs
B-type cyclins have HxKF signature
Cyclin destruction boxes (B1-type cyclin R-[AV]LGDIGN)
10](https://image.slidesharecdn.com/bitscompgenomics201102genefamilyanalysis-120116040620-phpapp02/85/BITS-Comparative-genomics-gene-family-analysis-10-320.jpg)
![Advanced methods for protein
(orthology) clustering
Sequence similarity-based
COG (RBH) [Tatusov 1997]
InParanoid [Remm et al., 2001]
Tribe-MCL [Van Dongen 2000]
OrthoMCL [Li et al., 2003]
Phylogenetic tree-based
PhylomeDB [Huerta-Cepas et al., 2007]
Ensembl Compara [Vilella et al., 2008]
24](https://image.slidesharecdn.com/bitscompgenomics201102genefamilyanalysis-120116040620-phpapp02/85/BITS-Comparative-genomics-gene-family-analysis-24-320.jpg)
BITS - Comparative genomics: gene family analysis
- 1. Comparative genomics in eukaryotes Gene family analysis Klaas Vandepoele, PhD Professor Ghent University Comparative & Integrative Genomics VIB – Ghent University, Belgium http://www.bits.vib.be
- 2. Workflow 2
- 3. Applications of clustering the proteome(s) Gene families form the basis for the evolutionary (or phylogenetic) analysis of Detection of orthologs and paralogs Gene duplication, family expansions, pseudogene formation and gene loss Species taxonomies Horizontal Gene Transfer (HGT) Evolution of gene structure • Introns • Protein domain organisation & (re)arrangements Base composition and codon usage 3
- 4. I. Structural annotation: genome- wide versus family-wise Rationale family-wise annotation Since every gene has different (sequence) characteristics and different genes evolve at different rates, using these characteristics to determine homologous gene models will improve the overall structural annotation quality Properties: Slow & nearly-manual procedure High-quality gene models revealing biological novel findings 4
- 5. Workflow family-wise annotation procedure Collecting experi- MSA experimental Family HMMbuild mental representatives representatives HMM profile EST/cDNA BLAST Species X proteome Protein motifs Ab initio gene prediction Correction gene model Putative HMMsearch Homologs Classification using Phylogenetic trees 5 Detailed characterization http://hmmer.janelia.org/
- 6. Experimental representatives InterProScan PFAM HMM logo Clustalw + JalView 6
- 7. BLAST / HMMsearch 1. Use multiple sequence alignment to create HMM profile 2. Use HMM profile to search for similar proteins 7
- 8. Representatives + putative homologs BioEdit Sequence Editor Suffix finalcds indicates corrected gene model compared to the original gene model generate by the ab-initio gene prediction Multiple sequence alignments assist in the detection and correction of errors in the structural annotation (missed exon) 8
- 9. Representatives + putative homologs Suffix finalcds indicates corrected gene model compared to the original gene model generate by the ab-initio gene prediction Multiple sequence alignments assist in the detection of errors in the structural annotation (false first exon) 9
- 10. Examples of family-specific protein motifs B-type cyclins have HxKF signature Cyclin destruction boxes (B1-type cyclin R-[AV]LGDIGN) 10
- 11. Examples of family-specific protein Arabidopsis Rice motifs D-type cyclins contain LxCxE Rb-binding motif Low conservation of phylogenetic signal at primary sequence level General rules are rarely general: exceptions (i.e. missing protein motifs) are frequent and might indicate functional divergence 11
- 12. Classification using phylogenetic tree construction A- and B-type cyclins are mitotic cyclins D-type cyclins are G1-specific H-type cyclins regulate activity of CDK-activating kinases • The complexity of the cyclin gene family appears to be higher in plants than in mammals • Whether there is functional redundancy within A- and B-type cyclins or different regulation (and expression) of some cyclin subclasses remains to be analyzed 12
- 13. Unraveling functional divergence using Genes large-scale expression compendia 13 Plant tissues
- 14. Unraveling functional divergence using large-scale expression compendia A-type cyclin B-type cyclin Genes D-type cyclin 14 Plant tissues Genevestigator
- 15. II. Orthology & paralogy A major goal of sequence analysis is evolutionary reconstruction. It is critical to distinguish between two principal types of homologous relationships, which differ in their evolutionary history and functional implications. Orthologs, defined as homologous genes evolved through speciation (~evolutionary counterparts derived from a single ancestral gene in the last common ancestor of the given two species) Paralogs, which are homologous genes evolved through duplication within the same (perhaps ancestral) genome. These definitions were first introduced by Fitch (1970) 15
- 16. Orthology & paralogy inference Organism phylogeny Gene phylogenies (species tree) gene duplication a1 A b1 B c1 a1 b) a2 a2 C b2 b1 c2 a) b2 speciation Outparalogs 16 Inparalogs c1
- 17. In- and outparalogy 17 Sonnhammer & Koonin: Orthology, paralogy and proposed classification for paralog subtypes
- 18. Tree reconciliation The automatic detection of speciation and duplication events using a species tree and gene family tree 18
- 19. III. Types of proteome analysis 19
- 20. The evolution of multi-domain proteins 20
- 21. Interpreting the output of an all- against-all similarity search Metrics for sequence similarity: • E-value, Bit score or percent identity 21 • alignment coverage
- 22. Clustering of similar sequences Proteins = vertices ~ nodes Sequence similarity relationship = edges 22
- 23. Clustering of similar sequences 23
- 24. Advanced methods for protein (orthology) clustering Sequence similarity-based COG (RBH) [Tatusov 1997] InParanoid [Remm et al., 2001] Tribe-MCL [Van Dongen 2000] OrthoMCL [Li et al., 2003] Phylogenetic tree-based PhylomeDB [Huerta-Cepas et al., 2007] Ensembl Compara [Vilella et al., 2008] 24
- 25. Overview methodologies BBH Inparanoid COG species overlap 25 Gabaldon, 2008 reconciliation
- 26. IV. Resources 26
- 27. Resources (bis) Ensembl (Vertebrates) EnsembGenomes (Metazoa, Protists, Fungi, Plants & Bacteria) OrthoMCLDB 5 (150 genomes) YGOB (>15 Fungi) 27
- 28. Hands-on Goal: identify and characterize gene family members encoding for talin 2 (TLN2) 1. Select Query gene 2. Retrieve homo/orthologs 3. Create multiple sequence alignment 4. Identify conserved positions 5. Create phylogenetic tree and identify ortho/paralogous genes 28