BioJS Human Genetic Variation Viewer

BioJavascript Human Genetic Variation Viewer
presented at the USC/UCLA Joint Bioinformatics Meeting held at University Of Southern California.

Developed as part of Google Summer of Code 2014

Transcript

1. human genetic variation viewer . Saket Choudhary 1, Leyla Garcia2 and Andrew Nightingale2 October 30, 2014 . . C . G . C . A . T . C . G . A . G . C . T . . C . G . C . G . T . C . G . A . G . C . T 1University of Southern California and 2EMBL-EBI
2. outline ∙ Motivation ∙ Solution ∙ Demo and Use-Cases ∙ Implementation ∙ Future Work 1
3. . .motivation
4. visualizations are powerful! The power of the unaided mind is highly overrated. The real powers come from devising external aids that enhance cognitive abilities. – Donald Norman 3
5. motivation ∙ NGS has given rise to catalog of genetic variants: dbSNP, COSMIC... 4
6. motivation ∙ NGS has given rise to catalog of genetic variants: dbSNP, COSMIC... ∙ Different categories of mutations: Benign, Damaging, Intermediate 4
7. motivation ∙ NGS has given rise to catalog of genetic variants: dbSNP, COSMIC... ∙ Different categories of mutations: Benign, Damaging, Intermediate ∙ Lack of consensus amongst scoring mechanisms: SIFT ̸= Polyphen 4
8. motivation ∙ NGS has given rise to catalog of genetic variants: dbSNP, COSMIC... ∙ Different categories of mutations: Benign, Damaging, Intermediate ∙ Lack of consensus amongst scoring mechanisms: SIFT ̸= Polyphen ∙ Lots of mutations =⇒ Loads of differing predictions 4
9. motivation ∙ NGS has given rise to catalog of genetic variants: dbSNP, COSMIC... ∙ Different categories of mutations: Benign, Damaging, Intermediate ∙ Lack of consensus amongst scoring mechanisms: SIFT ̸= Polyphen ∙ Lots of mutations =⇒ Loads of differing predictions ∙ Exploratory visualization is the first step towards discovering patterns, comparing consensus, aggregating predictions 4
10. motivation ∙ NGS has given rise to catalog of genetic variants: dbSNP, COSMIC... ∙ Different categories of mutations: Benign, Damaging, Intermediate ∙ Lack of consensus amongst scoring mechanisms: SIFT ̸= Polyphen ∙ Lots of mutations =⇒ Loads of differing predictions ∙ Exploratory visualization is the first step towards discovering patterns, comparing consensus, aggregating predictions ∙ Variation viewers are practically absent, those present provide limited flexibility 4
11. solution ∙ A graphical hub to present annotated variants from different sources 5
12. solution ∙ A graphical hub to present annotated variants from different sources ∙ Incremental levels of abstractions 5
13. solution ∙ A graphical hub to present annotated variants from different sources ∙ Incremental levels of abstractions ∙ Scalable and Interactive exploration on the web browser 5
14. overview Figure: Overview 6
15. zoomed view Figure: Zoomed View 7
16. Demo http://saketkc.github.io/biojs 8
17. . .details
18. implementation ∙ Written in javascript using the d3js library 10
19. implementation ∙ Written in javascript using the d3js library ∙ Deployed as a BioJS component 10
20. implementation ∙ Written in javascript using the d3js library ∙ Deployed as a BioJS component ∙ Flexible system with ability to capture and react to user-actions 10
21. why biojs ∙ BioJS is a javascript library for developing visualization of the biological data 11
22. why biojs ∙ BioJS is a javascript library for developing visualization of the biological data ∙ 11
23. why biojs Reusable components that can talk to each other 12
24. data input ∙ Pre-generated JSON files ∙ Current version uses files generated by an unpublished webservice at EBI ∙ Protein variants only 13
25. data input { ”id”:”P00533_variant226”, ”sourceIds”:[”COSM1090877”,”COSM1090879”], ”position”:541, ”wild_type”:”L”, ”mutation”:”I”, ”frequency”:0.0, ”polyphenPrediction”:”benign”, ”polyphenScore”:0.0, ”siftPrediction”:”tolerated”, ”siftScore”:0.86, ”somaticStatus”:1, ”consequenceTypes”:”missense variant”, ”cytogeneticBand”:”7p11.2”, ”genomicLocation”:”7:g.55229314C>A” } 14
26. features ∙ User defined scoring criteria 15
27. features ∙ User defined scoring criteria ∙ Different levels of abstractions, tooltips 15
28. features ∙ User defined scoring criteria ∙ Different levels of abstractions, tooltips ∙ Overview: Condensed information 15
29. features ∙ User defined scoring criteria ∙ Different levels of abstractions, tooltips ∙ Overview: Condensed information ∙ Zoomed View: All annotations 15
30. features ∙ User defined scoring criteria ∙ Different levels of abstractions, tooltips ∙ Overview: Condensed information ∙ Zoomed View: All annotations ∙ SIFT, Polyphen, .... 15
31. features ∙ User defined scoring criteria ∙ Different levels of abstractions, tooltips ∙ Overview: Condensed information ∙ Zoomed View: All annotations ∙ SIFT, Polyphen, .... ∙ Scalable, adaptable to new scores, mutation categories 15
32. use cases ∙ Identifying most or least mutated sites on a protein 16
33. use cases ∙ Identifying most or least mutated sites on a protein ∙ Discover differences between different scoring criteria 16
34. use cases ∙ Identifying most or least mutated sites on a protein ∙ Discover differences between different scoring criteria ∙ Benchmarking predictions 16
35. improvements ∙ VCF support(almost there!) 17
36. improvements ∙ VCF support(almost there!) ∙ Integration with Galaxy, web based bioinformatics workflows 17
37. improvements ∙ VCF support(almost there!) ∙ Integration with Galaxy, web based bioinformatics workflows ∙ Performance improvements 17
38. improvements ∙ VCF support(almost there!) ∙ Integration with Galaxy, web based bioinformatics workflows ∙ Performance improvements ∙ Interaction with 3D Protein viewer to highlight domains 17
39. . .conclusion
40. summary ∙ A tool for visualizing genetic variants ∙ Limited applications as a standalone tool, more usable with Protein Features Viewer ∙ Supports visualization of different levels of information ∙ Cross component talks ∙ User defined and user controlled ∙ Open Sourced(MIT License): https://github.com/saketkc/ biojs-genetic-variation-viewer 19
41. acknowledgements Google | Google Summer of Code 2014 BioJS Community Egor Dolzhenko USC MCB 20
42. Questions? 21