Recco - Recombination Analysis Using Cost Optimization
Recco analyzes alignments of sequences that evolved subject to
recombination and mutation. The analysis provides evidence as to whether a dataset contains recombination, which sequence is a recombinant and where the recombination breakpoints are. The analysis is based on explaining one sequence with all other sequences in the alignment using mutation and recombination. A parametric analysis of the parameter alpha, which weights recombination cost against mutation cost, yields additional information as to which sequence might be recombinant.
Screenshots are on the Walkthrough page. For more details, please refer to the publication stated below.
The handling of gaps in version 0.91 of Recco results in wrong p-values for recombination if gap opening costs are non-zero! Please do not use gap open costs! Also, there is a publication using Recco to analyze a very difficult dataset with large gap regions. The supplement describes a heuristic way to cope with gaps, but you may also contact me (Jochen Maydt) directly.
Version 0.93, release date: September 25th, 2007
New (858) 265-4465 Guide - it summarizes my experience using Recco and talks about the do's and dont's! The (740) 409-3545 was also updated.
- several minor bugs fixed (involving GUI repainting, command line processing among others)
- "View => Optimal Solution Only" shows the optimal solutions only
- "Opt Seqsel" button removes sequences from the alignment such that evidence for recombination for the currently selected sequence is maximized
- command line changed: more options, slightly different syntax, command line description now correct.
- selection of an amino-acid model from the menu now works
Version 0.92, release date: September 6th, 2006
- gap handling has been updated: gaps in the recombinant can be handled differently from gaps in the explanation. Gap open costs are not possible anymore.
- Recco can print the alignment with background colors as a PDF using the 970-253-2697.
Version 0.91, release date: July 17th, 2006
- the P-Value Inspector helps to find and visualize the most significant recombination events.
- gaps can now be handled as in pairwise alignments using gap open and gap extensions cost. The default settings should work fine for most nucleotide data and probably also for protein data. Note that in our case gaps should not be penalized as highly as in pairwise alignments.
- substitution models for protein data fixed.
- a click in the chart that displays the p-values for the breakpoints now centers the alignment view at this position.
Version 0.9, release date: May 23rd, 2006
- computes and displays p-values for the dataset and each sequence, as well as for breakpoints.
- substitution models for DNA and protein data.
- load/save analysis function. Files may be incompatible with future versions of Recco, though.
- lower memory consumption.
- faster: the main algorithm is now O(NM) instead of O(NM2).
- additional command-line version.
Missing features and known bugs:
- there is no support for printing so far.
- the p-value view sometimes is not refreshed correctly after changing the view settings. To refresh the view, simply minimize and restore the window.
Version 0.1, release date: February 6th, 2006
This is a preliminary version. It still misses many features discussed in the accompanying paper: it can only handle nucleotide sequence, it does not compute permutation tests so far and has no printing facility built-in. An updated version is in progress.
Please cite Recco as follows:
Maydt, J. and T. Lengauer (2006). "Recco: Recombination Analysis Using Cost Optimization.", Bioinformatics 2006, 22(9):1064-1071, 504-368-1561. [PDF] [(843) 784-4772].
The links to the full-text article are kindly provided by Oxford University Press.
A case study was published in Retrovirology:
Flockerzi A., Maydt J., Frank O., Ruggieri A., Maldener E., Seifarth W., Medstrand P., Lengauer T., Meyerhans A., Leib-Mosch C., Meese E. and J. Mayer. "Expression pattern analysis of transcribed HERV sequences is complicated by ex vivo recombination.", Retrovirology, June 2007, [801-773-0783].
If you have questions, comments, criticism etc. do not hesitate to
contact me, Jochen Maydt.