We will be offering an R workshop December 18-20, 2019. Learn more.
The pairwise.seqs command will calculate uncorrected pairwise distances between sequences. The command will generate a column-formatted distance matrix that is compatible with the cluster command. The command is also able to generate a phylip-formatted distance matrix. There are several options for how to handle gap comparisons and terminal gaps. This tutorial uses the data files in AmazonData.zip.
To run pairwise.seqs an file must be provided in fasta format. By default, the sequences are aligned to each other using the needleman alignment method with match, mismatch, gapopen and gapextend set to 1.0, -1.0, -2.0, -1.0, respectively. A gap is only penalized once, terminal gaps are penalized, all distances are calculated, and only one processor is used.
mothur > pairwise.seqs(fasta=amazon.fasta)
By default, pairwise.seqs will only count a string of gaps as a single gap and will penalize for terminal gaps
mothur has several ways of calculating a distance based on how gaps are treated. First, the default option - onegap - only counts a string of gaps as a single gap. For example:
SequenceA ATGCATGCATGC SequenceB ACGC---CATCC
Would have two mismatches and one gap. The length of the shorter sequence is 10 nt, since the gap is considered as a single position. Therefore the distance would be 3/10 or 0.30. This is the distance calculating method employed by Sogin et al. (1995). The logic behind this type of penalty is that a gap represents an insertion and it is likely that a gap of any length represents a single insertion. This can be called within mothur by the command:
mothur > pairwise.seqs(fasta=amazon.fasta, calc=onegap)
DNADIST actually ignores gaps. For example the two sequences above would have a distance of 2/9 or 0.2222. This type of distance calculation does not make much sense for sequences that are known to have a significant number of insertions. This can be used in mothur using the nogaps option. mothur can use this distance calculating method with the command:
mothur > pairwise.seqs(fasta=amazon.fasta, calc=nogaps)
A final option is to penalize each gap. For the two sequences above, the distance would be 5/12 or 0.4167. This can be used in mothur using the eachgap option. mothur can use this distance calculating method with the command:
mothur > pairwise.seqs(fasta=amazon.fasta, calc=eachgap)
There is some discussion over whether to penalize gaps that occur at the end of sequences. For example, consider the following sequences:
Sequence1 ATGCATGCATGC Sequence2 ---CAAGTA---
If terminal gaps are penalized, as is the default (and we are using the calc=onegap option), then the distance would be 4/8 or 0.5000. If they are not penalized, then the distance would be 2/6 or 0.3333. Ideally, all sequences would be aligned over the same region; however, if this is not possible or desired for some reason the ends option can be employed to tell mothur to ignore the penalization:
mothur > pairwise.seqs(fasta=amazon.fasta, countends=F)
The default is for countends to equal T.
If you know that you are not going to form OTUs with distances larger than 0.10, you can tell mothur to not save any distances larger than 0.10. This will significantly cut down on the amount of hard drive space required to store the matrix. This can be done as follows:
mothur > pairwise.seqs(fasta=amazon.fasta, cutoff=0.10)
Without setting cutoff to 0.10 this command would have generated 4560 distances (i.e. 96x95/2 = 4560). With the cutoff only 56 distances are saved. The savings can be substantial when there are a large number of distances. The actual cutoff used by the command is 0.005 higher than the value that is set to allow for rounding in the clustering steps.
The processors option will allow you to multiple processors to calculate a distance matrix. Default processors=Autodetect number of available processors and use all available. Multiple processors can be called as follows:
mothur > pairwise.seqs(fasta=amazon.fasta, cutoff=0.10, processors=2)
The output option allows you specify the form of the matrix generated by pairwise.seqs. By default, pairwise.seqs will generate a column-formatted matrix. You can set the output to "lt", for a phylip formatted lower triangle matrix, or to "square" for a phylip formatted square matrix. If output is set to lt or square the cutoff option is ignored.
mothur > pairwise.seqs(fasta=amazon.fasta, output=lt)
The pairwise.seqs command allows you to select between three alignment methods - blastn, gotoh, and needleman - needleman is the default setting:
mothur > pairwise.seqs(fasta=amazon.fasta, align=needleman)
The needleman algorithm penalizes the same amount for opening and extending a gap. Alternatively, you could use the gotoh algorithm, which charges a different penalty for opening (default=-2) and extending (default=-1) gaps:
mothur > pairwise.seqs(fasta=amazon.fasta, align=gotoh)
Our experience has shown that the added parameters in the gotoh algorithm do not improve the pairwise alignment and increases the time required for the alignment. Finally, blastn can be used as a heuristic approach to the gotoh alignment:
mothur > pairwise.seqs(fasta=amazon.fasta, align=blast)
In our implementation, blast is the slowest option of the three and also generates the worst alignments. The quality suffers particularly because it generates a local alignment (needleman and gotoh are global) and will truncate the alignment if the similarity drops below a threshold.
match, mismatch, gapopen, and gapextend
In the pairwise alignment portion of the aligning procedure, the default reward for a match is +1 and the penalties for a mismatch, opening and extending a gap are -1, -2, and -1. Our experience has shown that these produce the best alignments for 16S rRNA gene sequences. You are encouraged to play around with these to suit your own purposes as shown below:
mothur > pairwise.seqs(fasta=amazon.fasta, align=gotoh, match=1, mismatch=-3)
mothur > pairwise.seqs(fasta=amazon.fasta, align=gotoh, gapopen=-5)
Keep in mind that if you are using the align=blast option, blast will limit the combinations of match, mismatch, gapopen, and gapextend that you can use. Hopefully, we've scared you off of using blast at all so that this won't be an issue.
oldfasta & column
These parameters are used to append to a column-formatted distance matrix. Given a column matrix, a new fasta file and a old fasta file we add distances to the original distance matrix calculated with the old fasta file.
mothur > pairwise.seqs(oldfasta=final.fasta, column=final.dist, fasta=amazon.fasta)
- 1.24.0 - paralellized for Windows.
- 1.25.0 - added checks for positive values for gap open, gap extend, or mismatch.
- 1.30.1 - Bug Fix: Windows paralellization of pairwise.seqs cutoff was not passed to thread.
- 1.40.0 - Rewrite of threaded code. Default processors=Autodetect number of available processors and use all available.
- 1.41.0 - Added oldfasta and column parameters. #468