make.contigs alternative command

Commands in mothur
1

Hi there,

I’m totally newbie about using mothur and I have a question about the make.contigs command…

Well, it uses a fastq forward and reverse file to create a pairedend fasta file (after the we run the command). That’s ok, but what should I do when I have only one ffastq and rfastq and I have other samples to analyze, but they’re already in the fasta format?

If it’s too confuse, let me try to explain better:

My files are: sample16Sf.fastq and sample16Sr.fastq
anotherfastafile.fasta

So, how can I run mothur using only fasta files and not fastq files? I just need to change the mothur “stability.file” archive by adding those file names on it?

I’m a little bit lost and sorry for my newbie knowledge about this.

Thanks,
Rafael.

2

Probably the best way to do this would be to assemble the forward/reverse sequences then combine the output fasta file with anotherfastafile.fasta. for example;

mothur > make.contigs(ffastq=sample16Sf.fastq, rfastq=sample16Sr.fastq)
mothur > merge.files(input=sample16sf.contigs.trim.fasta-anotherfastafile.fasta, output=final.fasta)
mothur > make.group(fasta=sample16sf.contigs.trim.fasta-anotherfastafile.fasta, groups=Sample16S-AnotherFasta, output=final.groups)

Another way would be to just convert the sample16Sf.fastq to fasta format using fastq.info (and quality filter it with trim.seqs) then combine the files, but the first option is a better one.

3

Hi again,

It worked until I run the command screen.seqs after I’ve aligned all the samples with the Silva Database File.
After running screen.seqs in order to delimit the maximum length of my sequences, I ran summary.seqs command and I got, at the first line (in the minimum percent), the alignment starting at -1 and ending at -1 too.

I’m using the same type of data for all the samples (DNA) and I’m also using trimmed files. Following the MiSeq -SOP tutorial, all the files are in the first screening process (I’ve download those files before the screening analysis).

So, my problem is: how to run mothur with different metagenome samples? It isn’t clear for me, because it should work by the way I’m executing the commands.

I’m trying now to get the alignment from all the sequences inside the merged fasta, but I’m running out of solutions and got a lot of questions :frowning:

Thanks for you help,
Rafael.

4

Hm, how exactly are you running the alignment command? With the flip=T flag?

Also, how many sequences are in your pre-aligned and aligned files?

5

I’m sorry for taking too long for give you an answer.

The problems I’ve found are: when I followed the MiSeq guide all my sequences got removed before I make the classify.otu command. Also, when I tried to do the classify.otu command without align my sequences I received a warning that they aren’t with the same length (because I din’t align them, ok so far…).

I know that I don’t need to follow all the steps (like the Mock group, because I don’t have one) and I can use the commands without following the guide.

I have 12 metagenome samples and 2 of them run perfectly following the MiSeq - SOP guide. The 10 remaining samples I’ve found at the MG-Rast web-server. My question is: How can I make the taxonomic distribution, without aligning my sequences against the Silva Database and using only the RDP taxonomic data set? I just want to compare them using Mothur.

Thanks,
Rafael.

6

I’m pretty sure you can just run classify.seqs with your unique sequences against the database without worrying about alignment. Alignment-trimming your database does improve classification accuracy but it’s not critical to success.

From there you could either create phylotypes instead of OTUs, try the vsearch OTU clustering, or (ideally) try to work out exactly where the sequences are getting lost. It’s a bit tedious, but if you run your data through the MiSeq SOP again and do a summary.seqs after every step you should be able to pretty quickly track down where the sequence loss is coming from.

7

Thanks for you help!

I’m starting to get it. I ran classify.seqs only and it worked without crashing.
I’ll try to follow your advice and I’m sure that now I’ll be able to perform those analysis.

Thank you, thank you so much for your help, time and attention,
Rafael.