Robust sequence alignments are critical to interpreting the evolution and function of nucleic acid and protein sequences. However, accurately aligning multiple sequences is one of the fundamental problems in evolutionary biology and genomics. Even with the advent of powerful algorithms, alignments can often contain ambiguous regions. Should these regions be culled from the alignment or ‘fine-tuned’ further? Could opening up the world of tricky alignments to citizen scientists crack the problem?
Open-Phylo designed by Jérôme Waldispühl from McGill University, Canada, and colleagues may offer one solution. The results of this endeavour are published in a recent Genome Biology study. Open-Phylo is a freely accessible crowd-computing platform that seeks to harness the assistance of citizen scientists (in the form of online gamers) to attempt to solve the multiple sequence alignment problem. Crucially, their contribution to this holy grail, allows citizen scientists themselves to upload their own sequence data to Open-Phylo and select the portions of sequence they wish the crowd of gamers ‘playing the alignment’ to focus on.
The system has been tested against commonly used multiple sequence alignment programs such as Multiz, PRANK, Muscle and T-Coffee using sequences from three cancer genes encoding the tumour suppressor protein p53, breast cancer type 1 susceptibility protein (BRCA1) and retinoblastoma protein (RB1). In all cases Waldispühl and colleagues demonstrate that their human-powered crowd computer can improve the quality of the alignment generated by any one algorithm alone. Moreover, because Open-Phylo is directly administrated by the scientific community it enables anyone to enter the system and use crowds of gamers to assist in their bid for an optimal alignment.
Open-Phylo represents a new model for human-computing platforms, one that is powered by the public and open to the public.