Mutation Instability Filter (MIF) - HealthHackAu2014/HealthHack2014 GitHub Wiki
- Charles Galea (Scientist & Problem Owner)
- Yuri Feldman (Dev)
- Savant Krishna (Dev)
- Roslyn Lau (UX Design)
- Michael Walker (Scientist)
- Shiho Takagi (Dev)
- Marguerite Evans-Galea (Scientist)
Identifying genetic mutations responsible for familial disease is unnecessarily time-consuming, laborious and expensive. Current approaches typically provide 10-100 candidate genes, and eliminating the false positives requires difficult testing of individual corresponding proteins. Many bioinformaticians lack the technical expertise to streamline and speed up this process.
Create a filter for bioinformaticians to greatly reduce the number of genes that must be tested. Several web-based applications exist to identify protein structures and assess their stability.
MIF will take the amino acid sequence generated by the mutated gene:
- find the corresponding protein structures, and:
- access multiple web-based stability-testing algorithms to identify those most likely to cause disease.
Our solution will decimate the time and cost associated with identifying harmful genetic mutations. It will reduce the number of typically 100+ proteins that must be tested for harmful changes to a mere handful (2-3), greatly expediting research into genetic disease. As it identifies the protein structure responsible, it also indicates potential drug targets for medical treatment.
Links to datasets used.
-
Link to github repository
-
Link to running examples where appropriate
-
Link to youtube video
-
LINK TO PRESENTATION SLIDES --PENDING
Lovingly crafted in Ruby
Databases referenced
- [RCSB Protein Data Bank] (http://www.rcsb.org/pdb/home/home.do)
Protein stability algorithms
- [iStable] (http://predictor.nchu.edu.tw/istable/)
- [FoldX] (http://foldx.crg.es/)
- [Duet] (http://bioinformatics.ca/links_directory/tool/35345/duet)
- [ERIS] (http://dokhlab.unc.edu/tools/eris/)
- [POPMusic] (http://dezyme.com/login)
- [CUPSAT] (http://cupsat.tu-bs.de/)
- [iMutant] (http://gpcr2.biocomp.unibo.it/cgi/predictors/I-Mutant3.0/I-Mutant3.0.cgi)
- [AUTO-MUTE] (http://proteins.gmu.edu/automute/AUTO-MUTE_Stability_ddG.html)
What went well?
- Nice pre-defined data set
- Clear objective with examples of outputs at each stage of user flow supplied by Charles
- Good blend of complementary strengths across team. Team members tended to gravitate toward their niches.
- iStable had the optionality to grab results from other similar protein stability algorithm programs (PopMusic, CUPSTAT, iMutuant, AutoMutant and itself) for the one search input.
- Problem could be broken down into several processing stages that could be worked on independently of each other.
What would you have done better?
- More Ruby developers to speed up the process of data scraping in early stages of project
- Going to a whiteboard sooner to understand user flows and processes
- More time to develop a hi-fidelity user interface design
- Integrate with other web applications, e.g. Galaxy
- Identify potential protein targets for drug development.
Mutation Instability Filter or MIF 2014-2015. Excerpts, code and links may be used, provided that full and clear credit is given to the Mutation Instability Filter or MIF team (Charles Galea, Yuri Feldman, Roslyn Lau, Michael Walker, Savant Krishna, Shiho Takagi and Marguerite Evans-Galea) and Mutation Instability Filter or MIF with appropriate and specific direction to the original content.