Variathon challenge for predicting genotype variants

Genome-scale algorithmics research group at the Department of Computer Science co-organizes Variathon challenge, where high-throughput sequencing -based genotype variant calling methods are evaluated on accuracy and efficiency. This way of  predicting genotype variants is nowadays almost a routine task in bio(medicine) laboratories, and any improvements in accuracy and efficiency are of wide interest, especially in cancer research.  The goal of the challenge is to identify strengths and weaknesses of current approaches.

The challenge is organized now first time and is at pilot phase. From the experiences gained the challenge will be developed to more realistic direction with different analysis scenarios covered and using real read data with wet lab verification in addition to current simulated data. Also there will be an effort to develop the evaluation metrics.

Similar challenges have been arranged earlier in the close areas such as fragment assembly and RNA transcript expresssion prediction.

The students at the Dept. have the opportunity to take part in the light-weight version of the challenge during Project in Biological Sequence Analysis, period IV.  In fact, the project took place first time during Spring 2011, where the idea for Variathon challenge was born.

27.02.2013 - 09:49 Veli Mäkinen
22.02.2013 - 12:45 Veli Mäkinen