Variation and Functional Genomics

1. Variation and Functional Genomics

2. Overview • Genomic Diversity (SNPs) • Variations in the Ensembl Browser • Human genome • HapMap • Gen2Phen and EGA • A bit about Functional Genomics

3. Genomic Diversity SNPs (Single Nucleotide Polymorphisms) base pair substitutions InDels insertion/deletion (frameshifts) occur in 1 in every 300 bp (human) ~3 billion base pairs in mammalian genomes!

4. Single nucleotide polymorphisms (SNPs) • Polymorphism: a DNA variation in which each possible sequence is present in at least 1% of the population • Most polymorphisms (~90%) take the form of SNPs: variations that involve just one nucleotide

5. z z x x y y w w v v Origin of SNPs Mutation in individuals Selection of alleles Increase of the allele to a substantial population frequency w Fixation of the allele in a populations SNP Adapted from Bioinformatics for Geneticists, Eds Barnes and Gray

6. Functional Consequences

7. Studying variation – why? • SNPs can cause disease (SNP in clotting factor IX codes for a stop codon: haemophilia) • SNPs can increase disease risk (SNP in LDL receptor reduces efficiancy: high cholesterol) • SNPs can affect drug response (SNP in CYP2D8, a gene in the drug breakdown pathway in the liver, disrputs breakdown of debrisoquine, a treatment for high blood pressure.)

8. Studying variation – why? • Determine disease risk • Individualised medicine (pharmacogenomics) • Forensic studies • Biological markers • Hybridisation studies, marker-assisted breeding • Understanding Evolution

9. Practical Applications 9 of 25

10. SNPs in Ensembl Most SNPs imported from dbSNP (rs……): Imported data: alleles, flanking sequences, frequencies, …. Calculated data: position, synonymous status, peptide shift, …. For human also: HGVbase Affy GeneChip 100K and 500K Mapping Array Affy Genome-Wide SNP array 6.0 Ensembl-called SNPs (from Celera reads and Jim Watson’s and Craig Venter’s genomes) For mouse, rat, dog and chicken also: Sanger- and Ensembl-called SNPs (other strains / breeds) 10 of 25

11. dbSNP Central repository for simple genetic polymorphisms: single-base nucleotide substitutions small-scale multi-base deletions or insertions retroposable element insertions and microsatellite repeat variations http://www.ncbi.nlm.nih.gov/SNP/ For human (dbSNP build 129): 19,125,432 submissions (ss#’s) 2,920,818 new RefSNPs (rs#’s) 11 of 25

12. SNPs in Ensembl - Types Non-synonymous In coding sequence, resulting in an aa change Synonymous In coding sequence, not resulting in an aa change Frameshift In coding sequence, resulting in a frameshift Stop lost In coding sequence, resulting in the loss of a stop codon Stop gained In coding sequence, resulting in the gain of a stop codon Essential splice site In the first 2 or the last 2 basepairs of an intron Splice site 1-3 bps into an exon or 3-8 bps into an intron Upstream Within 5 kb upstream of the 5'-end of a transcript Regulatory region In regulatory region annotated by Ensembl 5' UTR In 5' UTR Intronic In intron 3' UTR In 3' UTR Downstream Within 5 kb downstream of the 3'-end of a transcript Intergenic More than 5 kb away from a transcript 12 of 25

13. Human • Chimp • Mouse • Rat • Dog • Cow • Platypus • Chicken • Zebrafish • Tetraodon • Mosquito SNPs in Ensembl - Species

14. Overview • Genomic Diversity (SNPs) • Variations in the Ensembl Browser • Human genome • HapMap • Gen2Phen and EGA • A bit about Functional Genomics

15. Focus on Human • Venter and Watson genomes • 1000 genomes project • HapMap

16. First diploid genomes for human Craig Venter: • Sequence & analysis ongoing since 2003 Jim Watson: • 454 technology (7.4x) • 100 mill unpaired reads (25 billion bps) • $1,000,000 “The Diploid Genome Sequence of an Individual Human” PLoS Biology 5: 10 2113-2144 (2007) “The Complete Genome of an Individual by Massively Parallel DNA Sequencing” Nature 452:872-876 (2008) “Accurate Whole Human Genome Sequencing Using Reversible Terminator Chemistry ” Nature 456:53-59 (2008) “The Diploid Genome Sequence of an Asian Individual” Nature 456:60-65 (2008)

17. www.1000genomes.org

18. 1000 Genomes • Delivering 20TB of sequence data… • First Pilot. 60 HapMap samples sequenced (low coverage) • Second Pilot. Two trios of European and African descent (high coverage) • Third Pilot. Sequence 1,000 genes in 1,000 individuals (high coverage)

19. 1000 Genomes Browser Main page • Built on Ensembl • Navigation on the left hand side • Options as drop down menus • Currently only includes human data • In the future comparative genomics information will be available • All pages link to Ensembl and UCSC

20. Spot the difference!

21. Reference Sequence • The Human Genome Project gave the “average” DNA sequence of a small number of people. • This helps us find out how a human develops and works • Does not show us the DNA differences between different humans • Does not reflect the major alleles

22. HapMap www.hapmap.org • A multi-country effort to identify and catalogue genetic similarities and differences in people. • Collaboration among scientists and funding agencies from Japan, the United Kingdom, Canada, China, Nigeria, and the United States. • All of the information generated by the project released into the public domain.

23. HapMap (phase I & II) • Samples from populations with African, Asian and European ancestry. • 270 DNA samples from 4 populations: • 30 trios (two parents and an adult child) from the Yoruba people of Ibadan, Nigeria • 45 unrelated Japanese from the Tokyo area • 45 unrelated Han Chinese from Beijing • 30 trios from Utah with Northern and Western European ancestry (CEPH)

24. HapMap (phase III) • Genotypes from 1115 individual from 11 populations: • ASW African ancestry in Southwest USA(71) • CEU Utah residents with Northern and Western European ancestry from the CEPH collection (162) • CHB Han Chinese in Beijing, China (70) • CHD Chinese in Metropolitan Denver, Colorado (70) • GIH Gujarati Indians in Houston, Texas (83) • JPT Japanese in Tokyo, Japan (82) • LWK Luhya in Webuye, Kenya (83) • MEX Mexican ancestry in Los Angeles, California (71) • MKK Maasai in Kinyawa, Kenya (171) • TSI Toscani in Italia (77) • YRI Yoruba in Ibadan, Nigeria (163)

25. Haplotyping • A haplotype is a set of SNPs (on average ~25 kb) found to be statistically associated on a single chromatid and which therefore tend to be inherited together over time. • Haplotyping involves grouping subjects by haplotypes.

26. Linkage Disequilibrium LD is the deviation from equilibrium, or random association. (i.e. in a population, two alleles are always inherited together, though they should undergo recombination some of the time.)

27. Measures of LD • D = P(AB) – P(A)P(B) • D ranges from – 0.25 to + 0.25 • D = 0 indicates linkage equilibrium • dependent on allele frequencies, therefore of little use • D’ = D / maximum possible value • D’ = 1 indicates perfect LD • estimates of D’ strongly inflated in small samples • r2 = D2 / P(A)P(B)P(a)P(b) • r2 = 1 indicates perfect LD • measure of choice • High LD, or perfect LD, shows high association of SNPs.

28. Linkage Disequilibrium LD values between two variants are displayed by means of inverted coloured triangles going from white (low LD) to red (high LD).

29. Tag SNPs define a haplotype Adapted from Nature 426, 6968: 789-796 (2003)

30. Tag SNPs • ‘Tag SNPs’ define the minimum SNP set to identify a haplotype. • r2 = 1 between 2 SNPs means 1 would be ‘redundant’ in the haplotype.

31. Locus specific databases (LSDB) • Databases that focus on one gene or one disease • e.g. p53, ABO, collagen • e.g. Albinism, cystic fibrosis, Alzheimer’s disease • User communities: • Research groups-disease and function driven • Clinicians – driven by genetic testing of patients

32. LSDBs • >700 on the Human Genome Variation Society website

33. LSDB examples

34. Why is it difficult to merge these data? • Historical reasons. LSDBs sometimes • Use sequences which do not start at Methionine • Use transcript coordinates not genomic • Use a different transcript for reporting mutations • Regularly changes with new assemblies/gene builds • It may contain minor alleles or rare alleles • It may be inaccurate • Missing genes (e.g. no α-haemoglobin - Thalasemia) • Mixture of sequences from different individuals

35. Ensembl and LRGs • Define an exchange format for LRGs with the NCBI • Create an LRG website • Create a pipeline for receiving the data and creating an LRG • Extend e! databases to store LRGs • Develop an API to query LRGs and associated annotation • Consult with the LSDBs to develop useful visualisation tools • Build displays for LRG data and annotation

36. Why is this important for Ensembl • Ensembl has traditionally focused on an infrastructure for molecular biologists • Needs to expand to provide support for more stable transcript sequences used for reporting mutations • It will give central databases access to patient variation, genotype, phenotype and disease data • This will improve our data resources

37. Advantages to LSDBs • LRGs in Ensembl gives LSDBs access to: • Genome annotation (including comparative, functional genomics and variation data) • Data integration with other variation resources (dbSNP, EGA, 1000 Genomes, NHGRI GWA catalogue) • Sequence search and data mining tools • A Perl API to query the data • A genome browser website for visualisation in genomic context and local context • Promotes discoverability of LSDBs • Data is mapped from one assembly to the next

38. EGA- Repository for genotype data • www.ebi.ac.uk/ega/

39. Variations Team Fiona Cunningham Yuan Chen Will McLaren

40. Functional Genomics (Wikipedia): Functional genomics is a field of molecular biology that attempts to make use of the vast wealth of data produced by genomic projects (such as genome sequencing projects) to describe gene (and protein) functions and interactions. In Ensembl: Regulatory build using ENCODE project information Promoters and Enhancers from CisRED and VISTA FlyReg features (for Drosophila)

41. ENCODE Encylopedia Of DNA Elements Where are the promoter, enhancer, and other regulatory regions of the human genome? Pilot project showed: Use chromatin accessibility and histone modification analysis to predict TSS 14 June 2007, Nature

42. Regulatory Build Uses CTCF and DNAse1 data from multiple cell types as “core features”. Overlapping methylation sites expand these regions.

43. How to get there?

44. Click on a Regulatory Feature…

45. Region in Detail

46. BioMart

47. There are other sets… Sequence motifs determined by experimental and prediction tools. http://www.cisred.org/ VISTA Enhancer Set Tissue-specific enhancers. Tested experimentally. Nucleic Acids Res. 2007 January; 35(Database issue): D88–D92.

48. Total List of Regulation Info. • Homo sapiens • Mus musculus • Danio rerio • Drosophila melanogaster • DNase I Hypersensitivitiy sites for GM06990 and CD4+ T cells • CTCF binding sites • Histone modification data • MeDIP-chip methylation data for 17 human tissues and cell lines • VISTA Enhancer Assay (http://enhancer.lbl.gov) • cisRED motifs (www.cisred.org) • miRanda microRNA target prediction • Expression Quantitative Trait Loci (eQTL) from the Sanger Institute • DNase1 Hypersensititvity site (ES cells) • Histone modifications for ES, MEF, and NPC cells • cisRED motifs (www.cisred.org) • ZFMODELS-enhancers • REDfly TFBSs • BioTIFFIN • REDfly CRMs

49. Functional Genomics Team • eFG Ian Dunham • Nathan Johnson • Daniel Sobral (starts Dec 1) • Andy Yates (multi-species support) • ENCODE • Steven Wilder • Damian Keefe

50. End of course survey! http://tinyurl.com/yaw6nzq