Old version of mothur

Dear Pat,

is there a possibility to download mothur version 1.24.0. I have problems with the new version and cannot find the problem, however it run perfectly on the mentioned older version, which I do not have on my server.

Thanx a lot in advance,
B

Uhhh, we we’d like to know what the problem is. Can you tell us more about what is going on?

You can find all of the old source code at http://www.mothur.org/wiki/Download_mothur.

Pat

First of all, thanx a lot for your fast reply :slight_smile: !

My script looks as follows:
sffinfo(sff=in.sff, flow=T)
shhh.flows(file=in.flow.files, processors=8)
trim.flows(flow=in.flow, oligos=oligos.txt, bdiffs=1, pdiffs=2, processors=8, minflows=360, maxflows=720)
shhh.flows(file=in.flow.files, processors=8)
trim.seqs(fasta=in.shhh.fasta, name=in.shhh.names, oligos=oligos.txt, pdiffs=2, bdiffs=1, maxhomop=8, minlength=200, processors=8)
summary.seqs(fasta=in.shhh.trim.fasta, name=in.shhh.trim.names)
unique.seqs(fasta=in.shhh.trim.fasta, name=in.shhh.trim.names)
summary.seqs(fasta=in.shhh.trim.unique.fasta, name=in.shhh.trim.names)
align.seqs(fasta=in.shhh.trim.unique.fasta, reference=silva.bacteria.fasta, processors=8)
summary.seqs(fasta=in.shhh.trim.unique.align, name=in.shhh.trim.names)
screen.seqs(fasta=in.shhh.trim.unique.align, name=in.shhh.trim.names, group=in.shhh.groups, optimize=start, minlength=200, processors=8)
summary.seqs(fasta=in.shhh.trim.unique.good.align, name=in.shhh.trim.good.names)
count.groups(group=in.shhh.good.groups)
filter.seqs(fasta=in.shhh.trim.unique.good.align, vertical=T, trump=., processors=8)
unique.seqs(fasta=in.shhh.trim.unique.good.filter.fasta, name=in.shhh.trim.good.names)
pre.cluster(fasta=in.shhh.trim.unique.good.filter.unique.fasta, name=in.shhh.trim.unique.good.filter.names, group=in.shhh.good.groups, diffs=2)

after the pre.cluster command I get following output:
missing name HI4A5R103F02KF
missing name HI4A5R103F02OG
missing name HI4A5R103F0379
missing name HI4A5R103F03U6
missing name HI4A5R103F03WF
missing name HI4A5R103F049H
missing name HI4A5R103F04K1
missing name HI4A5R103F04N4
etc. (feels like 1000 pages of “missing name…”)

HOWEVER, looking for the error, I saw that the problem starts before, namely at:
shhh.flows(file=in.flow.files, processors=8)
-> Instead of getting the output for 27 samples, 3 samples are missing.
I thought I could simply take the same command, but using the single flow files of the three missing samples:
shhh.flows(flow=in.A5.flow, processors=8)
-> But the output was: segmentation fault (core dumped)
Does this mean there is not enough memory?

As mentioned I have applied my script on the same AND another set of samples (but also 16S rRNA gene amplicons) and it worked perfectly on mother 1.24.1. We are thankfully using the server of Thomas Rattei in Vienna, but since the update mothur regularly, 1.24.1 is no longer available and I tried to download the old version to my pc, which did not work.
So, now I will try your link (keeping my fingers crossed!) :roll:

I will inform you about the outcome.
Best regards,
B

Hi Brigitte,

I believe in 1.25 the output for unique.seqs has been changed (silent patch?). It now outputs XXX.unique.names, instead of XXX.names. So there s an error in your code:

unique.seqs(fasta=in.shhh.trim.fasta, name=in.shhh.trim.names)
summary.seqs(fasta=in.shhh.trim.unique.fasta, name=in.shhh.trim.names)
align.seqs(fasta=in.shhh.trim.unique.fasta, reference=silva.bacteria.fasta, processors=8)
summary.seqs(fasta=in.shhh.trim.unique.align, name=in.shhh.trim.names)
screen.seqs(fasta=in.shhh.trim.unique.align, name=in.shhh.trim.names, group=in.shhh.groups, optimize=start, minlength=200, processors=8)

Here, screen.seqs took wrong .names file, that is why you have missing data. I dont know about the 3 missing samples …

I am going to create new thread in bug section about unique.seqs. http://www.mothur.org/forum/viewtopic.php?f=4&t=1601

EDIT: Also, I would suggest you to use a criteria= parameter with optimize=, unless there s a default value.

It worked!!! :slight_smile:

unfortunately I still miss the 3 samples, but I continue to work on that…

Thanx a lot for you help!

When I took a closer look at your 1.25 logfile, I think I found the source of the problem. In the shhh.flows command, there is no output for the processing of (file 5 of 27) or (file 6 of 27). I think what is happening is that the process that was running those files crashed somewhere in file 4. This is most likely due to a memory limitation. The change you made to the min and max flows in trim.flows most likely increased the size of the flow files sent to shhh.flows. It is strange that the crashing process did not report an error message before dying. I am going to add some extra checks to mothur, so that in the future we will make sure all the spawned processes completed all the files they were sent. I would suggest running shhh.flows with 1 processor as a workaround in the meantime. The more processors you use the more memory is required, so running the A5, A7, and A8 flow files without any other files being processed at the same time may help. Thanks for bringing this issue to our attention!

Hi Mothur-users,

I’m having the very same problem with a sample. I’m using Mothur v.1.33.3 and when I try running shhh.flow on this specific sample, it crashes after identifying the unique flowgrams.

Denoising flowgrams…
iter maxDelta nLL cycletime
segmentation fault (core dumped)


I did it with 1 processor as well and the result is the same. I tried on a multiple CPU server but it doesn't work either... Do you have any idea what I could do to solve this ?

Thanks very much in advance!

Thomas

Thomas,

The most likely problem is that your memory is getting swamped out. How are you running trim.flows? Can you post the output from running shhh.flows?

pat