Current version: 1.3

Cogent is a tool for reconstructing the coding genome using high-quality full-length transcriptome sequences. It is designed to be used on Iso-Seq data and in cases where there is no reference genome or the ref genome is highly incomplete.

See a recent presentation on Cogent being applied to the Cuttlefish Iso-Seq data.

A preliminary draft paper is available. Note that there have been minor implementation changes since the draft was written, however the concept and overall methodology remains.

Please see wiki for details on usage.

2016.03.02 updated to version 1.1. Found several cases where border cases caused program crash. Fixed.

2016.03.21 updated to version 1.2. Fixed more minor bugs related to edge cases esp caused by non-obvious graph cycles.

2016.03.30 updated to version 1.3. Important bug in splice_graph.contract_sinks fixed where it previously was accidentally contracting sinks with predecessors that had multiple outgoing edges, which causes incorrect reconstructions in cases where there are a lot of isoforms with alternative 3' ends.