My project involves whole genome alignment and analysis of relatively small, highly homologous prokaryotic genomes. I have used both MAFFT and MAUVE to align the genomes and whilst I have noticed some small differences, for example - how the aligner handles gaps/insertions in sequences, the output produced by both algorithms is essentially identical. I understand if (and when) I require alignment of more divergent genomes that MAUVE would be a better solution, but in the case of small, homologous genomes does it really matter which one I use?
Hi Eileen,
In the case of small highly similar genomes both MAFFT and MAUVE work fine. So it does not really matter which one you use. An important issue with aligning nucleotide sequences--short or long--is to choose the correct or appropriate parameters (mainly scoring matrix and gap opening/extension penalties). Also, when genomes are to be aligned visualization of rearrangements, genes, and conservation levels is very useful. Gene annotation also. In this respect MAUVE is superior to MAFFT.
And a note on homology: one of my mentors used to say that homology is like pregnancy, either you are pregnant or not. Similarly with homology, either you are homologous,that is you share common ancestry, or not. Sequence similarity can be high or low.
Hope this helps a bit.
Dimitra
Hi Eileen,
In the case of small highly similar genomes both MAFFT and MAUVE work fine. So it does not really matter which one you use. An important issue with aligning nucleotide sequences--short or long--is to choose the correct or appropriate parameters (mainly scoring matrix and gap opening/extension penalties). Also, when genomes are to be aligned visualization of rearrangements, genes, and conservation levels is very useful. Gene annotation also. In this respect MAUVE is superior to MAFFT.
And a note on homology: one of my mentors used to say that homology is like pregnancy, either you are pregnant or not. Similarly with homology, either you are homologous,that is you share common ancestry, or not. Sequence similarity can be high or low.
Hope this helps a bit.
Dimitra
Hi Eileen:
If you are going to choose only one, I would recommend MAUVE as it has become the "industry standard" and I agree with Dimitra that the visualization tools are better. This is not to say that MAFFT is algorithmically inferior and in fact we routinely use both because as with any research you want to be able to validate your findings, but when building figures for publication we tend to use MAUVE
Hi Eileen,
I am not aware of any systematic comparison between MAUVE and MAFFT, but know the performance of MAFFT is very good compared to several other multiple sequence alignment (MSA) methods (Mol Biol Evol 24, 2433-2442; Bioinformatics 28, 495-502). Some of the commonly-used MSA methods tend to align the gaps (Mol Biol Evol 24, 2433-2442) rather than the amino acids, which is a particular problem when trying to align alternatively-spliced gene products. This might not be a problem in your case, though.
All the best,
Lars
I usually use MAUVE for large genome comparisons (range of Mb) and MAFFT for single proteins (or cancatenate) alignments. Mauve allow you to identify "co-linear" blocks of shared similarity but I also agree with Dimitra that tuning parameters can bring you to think that two genes are homologous or not depending by the threshold you use and this is not biologically meaningful.
maafft doesn't accept genomes in multiple contigs only one multifasta= complete genomes.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics
- Bioinformatic Tools