Congratulations to Deepank for his
new paper in Bioinformatics describing the first ever benchmark for gene finding. Deepank found that current gene finding programs are at least 5% inaccurate, mostly in their prediction of the correct start site.