make.contigs skips some groups in 1.34.4 version?

mothur bugs
1

Hi, I was wondering about the difference in the output of make.contigs command between 1.34.4 version of mothur and some older version, which used my colleague on the same fastq files.
More particularly, the command was:
make.contigs(ffastq=665fITS7_R2.fastq, rfastq=665fITS7_R1.fastq, oligos=ITS7ja5.oligos, pdiffs=2)

The output of my colleague’s run:
Group count:
fITS7R1_ITS4F1 11867
fITS7R1_ITS4F2 11547
fITS7R1_ITS4F3 9927
fITS7R2_ITS4F1 17167
fITS7R2_ITS4F2 16679
fITS7R2_ITS4F3 14482
fITS7R3_ITS4F1 10056
fITS7R3_ITS4F2 9772
fITS7R3_ITS4F3 8486
Total of all groups is 109983

While my output contains less groups and total number of groups:

Group count:
fITS7R1_ITS4F1 3850
fITS7R1_ITS4F2 3573
Total of all groups is 7423

When I tried to relax pdiffs from 2 up to 8, the total of all groups increased, but the number of groups which passed into the output was only four.

The only difference compared to colleague’s run was, that I made small modification of oligos file by adding the line indicating the absence of barcodes in reads after primers, which was obligatory for 1.34.4 version.

BARCODE NONE NONE q


So is it some kind of bug or changed algorithm in current version compared to older?
Thank you in advance for any suggestions.

PS sorry that I did not specify the older mothur version; the colleague is not available now for some time.

2

Sorry, but it really depends on which version of mothur your colleague is using. Some older versions had make.contigs in it before we developed the method in the Kozich paper although we didn’t announce the command. This is one of the reasons it’s great to use the latest version.

Pat

3

Thanks for answering,
the version of my colleague mothur is v.1.33.0 with the last update 4/4/2014, is it the version with make.contigs before the method in the Kozich paper?
In other words, could I really more confidently on new version, and it is not some kind of a bug?

4

Before version 1.34.0, mothur allowed pdiffs differences in the forward primer and pdiffs differences in the reverse primer. Similarly, mothur allowed bdiffs differences in the forward barcode and bdiffs differences in the reverse barcode. So if pdiffs=2, you could have 2 differences in the forward and 2 diffs in the reverse. In version 1.34.0, we made a change so that the pdiffs allowed = the diffs in the forward + the diffs in the reverse. So if pdiffs=2, you can have a total of 2 differences in paired primer, i.e.(2 in the forward and 0 in the reverse, 1 in the forward and 1 in the reverse, or 0 in the forward and 2 in the reverse).