Theoretical Computer Science and Algorithmic Methods



COLOMBO is a software framework equipped with a GUI for the prediction of genomic islands (GIs) in prokaryotes. It can be equipped with different plugins that actually perform the prediction. The current version of COLOMBO is supplied with

- SIGI-HMM that uses hidden Markov models and codon usage bias;

- SIGi-CRF utilizing conditional random fields and tetramer-based order-3 Markov chain transition probabilities. In particular, SIGI-CRF analyzes non-annotated genomic sequences.


SIGI_CRF-AUTO is an alternative front-end to the SIGI-CRF processing core. First, it allows to batch process a large number of genome sequences and/or to run a variety of parameter combinations. Second, it can collect all individual prediction results generated by varying parameter combinations and combine them into a single graphical representation. Once rendered, images can be saved very flexible with a broad variety of format options.

Look in the download section to download the current versions of COLOMBO and SIGI_CRF-AUTO for unix (tar.gz file) and for windows (zip file).