This page contains information related to my work on the evolution of eukaryotic gene structure (exon-intron organization).

Software: Malin (2008-)

Malin is a free software package for the analysis of eukaryotic intron evolution [announced by an Application Note in Bioinformatics]. See Malin's website.

99+1 eukaryotes (2011)

With Igor Rogozin and Eugene Koonin, we reconstructed history using Markov Chain Mointe Carlo [PLoS Computational Biology]. See the companion site.

Intron-rich alveolate ancestors (2008)

My paper entitled Extremely intron-rich genes in the alveolate ancestors inferred with a flexible likelihood approach, written with Igor Rogozin and Eugene Koonin [Molecular Biology and Evolution], discusses the application of a rates-across-sites model to chromalveolate intron evolution. For further information, see the companion site.

In search of lost introns (2007)

In search of lost introns was written with Andrew Holey and Igor Rogozin, and presented at the ISMB/ECCB conference in 2007. You can download the paper from the journal Bioinformatics, or a preprint from arXiv.

Software downloads

There are two Java packages you can download. The first one is IntronAlignment.jar, for (re-)alignment of intron-annotated protein sequences, and the second one is intronRates.jar, for optimizing the likelihood and computing histories. Instructions on how to use the programs are in the manuals: intronAlignment-doc.pdf and intronRates-doc.pdf. Please send me an e-mail about your experience if you use the programs.

What is in the packages?


The method was applied to a newly compiled data set of 18 eukaryotes: 8044 orthologous intron positions in 483 genes. The species are the following: Homo sapiens (Hsap), Rattus norvegicus (Rnor), Takifugu rubripes (Trub), Danio rerio (Drer), Drosophila melanogaster (Dmel), Anopheles gambiae (Agam), Apis mellifera (Amel), Caenorhabditis elegans (Cele), Caenorhabditis briggsae (Cbri), Saccharomyces cerevisiae (Scer), Neurospora crassa OR74 (Ncra), Schizosaccharomyces pombe 972h- (Spom), Ustilago maydis 521 (Umay), Cryptococcus neoformans v. n. JEC21 (Cneo), Oryza sativa ssp. japonica (Osat), Arabidopsis thaliana (Atha), Plasmodium falciparum 3D7 (Pfal), Plasmodium berghei str. ANKA (Pber). The picture below shows their phylogeny. Double lines show the few edges with significant net intron gain. Framed terminal taxa appear in the data set of Rogozin et al. (2003).

phylogeny of 18 species

The following pictures show the inferred intron density at ancestral nodes (error bars denote 95% confidence intervals), and gains/losses on branches (empty rectangles for the total gain and loss, and filled rectangles for net change; tree is rooted at node 16). Notice the high intron density at ancestral nodes, and the prevalence of loss.

intron counts   intron count changes

Likely scenarios (2005)

My initial ideas about reconstructing intron evolution were outlined in the paper Likely scenarios of intron evolution, presented at the Third RECOMB Satellite Workshop on Comparative Genomics (Springer LNBI 3678, pp. 47-60, 2005). You can download the preprint here: mle-introns.pdf.


I wrote a Java package to perform phylogenetic analysis of a presence-absence data set of introns in homologous positions. You can download the package here: intronLoss.jar. Instructions on how to use it are in the manual: intronLoss-doc.pdf. Please send me an e-mail about your experience if you use the programs. If you are really curious, I can send you the source code too.

What is in the package?


The method was applied to a data set compiled by Igor Rogozin, consisting of 7236 homologous intron sites in 8 Eukaryotes. The picture below shows the intron counts at the terminal taxa, and ancestral intron density based on likelihood optimization and posterior calculations. Branches are colored according to whether the number of gains is at least twice or at most half as large as the number of losses on them.

intron evolution in eukaryotes

It turns out that