Jakub Kovac, Tomas Vinar, Brona Brejova.
Predicting Gene Structures from Multiple RT-PCR Tests.
In S.L. Salzberg, T. Warnow, ed.,
Workshop on Algorithms in Bioinformatics (WABI 2009),
5724 volume of Lecture Notes in Bioinformatics,
pp. 181-193, Philadelphia, PA, USA, September 2009. Springer.
Preprint, 325Kb | Download from publisher | BibTeX
It has been demonstrated that the use of additional information such as ESTs and protein homology can significantly improve accuracy of gene prediction. However, many sources of external information are still being omitted from consideration. Here, we investigate the use of product lengths from RT-PCR experiments in gene finding. We present hardness results and practical algorithms for several variants of the problem. We also apply our methods to a real RT-PCR data set in the Drosophila genome. We conclude that the use of RT-PCR data sets can improve the sensitivity of gene prediction and locate novel splicing variants.