The coding capacity of the genome is greatly expanded by the process of alternative splicing, which enables a single gene to produce more than one distinct protein. Can the expression of these different proteins be predicted from sequence data? Here, modelling based on information theory has been used to develop a 'splicing code', which can predict, with good accuracy, tissue-dependent changes in alternative splicing.
- Yoseph Barash
- John A. Calarco
- Brendan J. Frey