Crowdsourcing gene predictions & estimating population sizes



Crowdsourcing gene predictions & estimating population sizes

0 0


seminar14

Talk given in a seminar at the Queen Mary, University of London.

On Github bmpvieira / seminar14

Crowdsourcing gene predictions & estimating population sizes

bmpvieira.com/seminar14

Bruno Vieira | @bmpvieira

Bioinformatics & Population Genomics

Initially address two issues

Scaling up gene prediction

Infer the efective population size history in insects with the PSMC method (Li, 2011).

Gene prediction?

Why is this important?

Genes are the basic building block of organisms

How?

Gene prediction models (Sleator, 2010)

Web application to crowdsource gene prediction

github.com/yeban/afra

Crowdsource?

Crowd + Outsource

Citizen Science

James Borrell | @James_Borrell Citizen Cyberscience Summit 2014 | #ccs14

Self-reward helping Science

Zooniverse success

Science? I don't care...

Cognitive surplus

Shirky, 2010

Gamification

Gamification

A way to engage users into solving a problem by adding game mechanics to it

Useless game - Flappy bird

50 milion downloads

flappybird.io

Useful - Genes In Space

cancerresearchuk.org

Previous work

Scale up and Gamify another Open Source project

gmod/apolloyeban/afra

Anurag Priyam | @yeban

Current work

Scale up

Move most of the logic to the browser

Scale up

Biology logic on the browser

github.com/bionode/bionode

Gamification

Dashboad mockup

Machine Learning

Use data generated by users to improve gene prediction models

Robert Simpson | @orbitingfrog Citizen Cyberscience Summit 2014 | #ccs14

PSMC

Effective population size?

Theoretical number of individuals that contribute gametes to the next generation

Why is this important?

Measure of genetic diversity

Affects selection efficiency

Used

Effect of historical climate changes (Miller, 2012)

Measure the impact of anthropogenic activity (Zhao, 2013)

Discover unexpected population bottlenecks (Freedman, 2014)

Detect the time of divergence between populations (Li, 2011)

How to measure?

Previously hard to do

  • Highly stochastic nature of inbreeding and genetic drift
  • Other confounding factors
  • Needs a lot of specific data

Now from a diploid genome

PSMC

Li, 2011

Hasn't been used in insects a lot... until now!

Use PSMC to answer some evolutionary questions

Is the effective population size in solitary insects > social?

Experimental design

Run PSMC across a wide range of social insects and their solitary relatives

Current work

Reproducing published results to master PSMC

Thank you!

Bruno Vieira | @bmpvieira

Anurag Priyam | @yeban

Yannick Wurm | @yannick__

bmpvieira.com/seminar14

© 2014 Bruno Vieira CC-BY 4.0

Crowdsource gene prediction
  • Address data "deluge" in gene prediction
  • Scale up by moving logics to browser
  • Gamify to tap into Cognitive Surplus
Effective pop. size history in insects
  • Deploy the PSMC on the servers
  • Master PSMC by reproducing results
  • Effective pop. size solitary insects > social?