Using sff tools to Extract Read Subsets

[Xterra]

How can I extract a subset of reads (longer than 150 bases but shorter than 200) from a sequencing Run using the sfftools. I think this can be accomplished using either sffinfo or fnafile (-t) but I just cannot get it to work.
Any help will be very much appreciated.

[sklages]

This is not directly possible with sfftools. You'll probably need to work on fasta files.

cheers,
Sven

[kmcarr]

You can't do this with sfftools directly. You can use sfffile to select a subset of reads from an input SFF file by using the -i option to pass a file containing a list of accession number to include in the output. Of course you will have to use some other tools to first create the list of accession numbers for the reads which match your criteria, whatever they may be.

(ETA: Oh, Sven beat me to it.)

[Xterra]

How about fnafile?

Could someone please explain me how to use -t option with fnafile? I mean if I use -i accno 1-200, wouldn't I be able to 'extract' the reads that are 150 bases long?

[kmcarr]

No, the -t option is used for changing trim point information stored in the file, not for extracting specific reads.

[Xterra]

fnafile -t syntax

Hmmm! Ok, I understand the limitations. Still, would you mind explaining me the syntax for fnafile using the -t option (accno 1 200)? I just want to see what that tool is capable to do and see if I can use it down the road.
Thanks!

[sklages]

fnafile --help

I never used this program, it seems to be the fasta counterpart to sffinfo.

Both sffinfo and fnafile are capable of filtering by read names (ids) via '-e' or '-i', they are not capable of filtering by any other characteristic (e.g. length, gc content etc.).

-t / -tr *set* trim points according to a file given by the user. These are not filter.

[Xterra]

Skalages

I just cannot get it to work. Would you mind being a little more specific? Let's assume I have the file A.fas which is a FASTA file and I am using 1 200 instead of 12 543 as described on the manual

The specifi ed “trimfi le” should contain one or more lines
consisting of (1) a read accession number, (2) a starting trimpoint
and (3) an ending trimpoint, separated by whitespace characters or
where the trimpoints are separate by a dash (e.g., “accno 12 543” or
“accno 12-543”)

[kmcarr]

I just cannot get it to work. Would you mind being a little more specific? Let's assume I have the file A.fas which is a FASTA file and I am using 1 200 instead of 12 543 as described on the manual

Could you please post the exact command you were using, as well as small samples of the FASTA and trim files you tried to use.

[Xterra]

I have tried so many different commands

but none is working. That's why I would like to get the right syntax.
I have uploaded an example of the FASTA file I would like to process using fnafile. As I said, I am only trying to find out what fnafile can do.
Thanks.

[kmcarr]

but none is working. That's why I would like to get the right syntax.

O.K., how about an example of just one command that didn't work.

And your FASTA file looks like it contains gapped sequence.

>GF2FOAC04ISOQO
ACGAG-TG----GTGATGT-GCCAGC-TG-CCGTTGGTGT-TAATGAGCTGAA-TGTTCT
GCTGA-G-------GGC--ATGGC-T-GAACAC-GACGG-CAAATCACGT----TGTGAA
CGTG-CAA-CACGCG-CC--TCAA-CGGT-GGTGGT-G--CCCG-CGT--CCACCCCA-G
CGG-CCAG-C-AGAAGGA--TGA-CAAT-GACCCTT-C--G-CCCACGACT---------
>GF2FOAC04J0H2I
ACGAA-TGCG-TTTGATGT-GCCAGC-TG-CCGTTGGTGT-TAATGAGCTGAA-TGTTCT
GCTGA-G---G---GCCGAGTGGCGTAGAACAC-GCCGG-CAAT-CA-GT-TGGTGG-AA
CGTG-CAA-CA-GCG-CC-TCCAA-C----GGGGGTCG--CCCG-CGC--CCACCCCA-G
CGG-CCAG-C-AGAAGGA--TGA-CAAT-GA-CCTT-C--G-CCCA--------------

Where did this FASTA come from?

[Xterra]

Here you have one without gaps.

Does it matter if you have gaps in the fasta file? Would it affect the performance of fnafile? Anyways, I have uploaded a file with no gaps. That's a 454 run using the Titanium amplicon sequencing kit. The previous file was the aligned file and that's why you could see the indels.

[sklages]

Does it matter if you have gaps in the fasta file? Would it affect the performance of fnafile? Anyways, I have uploaded a file with no gaps. That's a 454 run using the Titanium amplicon sequencing kit. The previous file was the aligned file and that's why you could see the indels.

You still haven't provided the exact command you used.

Your fasta file contains aligned sequences created by some "windows" program? OK, .. you provided a DOS formatted file, no idea if fnafile is happy about that.

1) create a file with trim points, just like:

GF2FOAC04I8T0F 1 200
GF2FOAC04J305H 1 200
GF2FOAC04J3QXL 1 200

2) run fnafile

$ fnafile -o out.fa -tr tp.txt UniqueHapsUnix.fas

3) see the differences

a) original sequence:

>GF2FOAC04I8T0F
ACGAGTGCGTTTGATGTGCCAGCTGCCGTTGGTGTTAATGAGCTGAATGTTCTGCTGAGGGCCATGGCTG
AACACGCCGGCAATCACGTTGGTGGAACGTGCAACAGCGCCTCCAACGGTGGTGGTGCCCGCGTCCACCC
CAGCGGCCAGCAGAAGGATGACAATGACCTTCGCCCAC

b) "trimmed" sequence

>GF2FOAC04I8T0F trim=1-200
ACGAGTGCGTTTGATGTGCCAGCTGCCGTTGGTGTTAATGAGCTGAATGTTCTGCTGAGG
GCCATGGCTGAACACGCCGGCAATCACGTTGGTGGAACGTGCAACAGCGCCTCCAACGGT
GGTGGTGCCCGCGTCCACCCCAGCGGCCAGCAGAAGGATGACAATGACCTTCGCCCAC

The sequence itself remains unchanged, but there has been a flag introduced (trim=1-200) directing the assembler (newbler) to just use the sequence within this range. Again, this not a filter and no physical trimming of the reads. You need to use an external tool to either work on fasta or sff files to trim your sequences.

[Xterra]

Thanks!

That's the answer I was looking for!

[TheLight]

SFF editor/convert

There is a free tool with graphic interface that allows you to view/edit and convert SFF files here