[mmartin]

How would the masking look like? I imagine lowercase vs. uppercase would be an option. Could you please add this as a feature request to the issue tracker on http://cutadapt.googlecode.com/ ?

[lmilne]

Masking with 'N' would work for me.

[kbhit]

What's the best way to cutadapt Illumina reads where there are 2 adaptors such as:
5' - adaptor1 - sequence - adaptor2 - 3'

I want to the adaptors on both sides. What I have been doing is running cutadapt twice:
a. Run cut-adapt with -a flag with adaptor2
b. Feed the above output to another cut-adapt with -g flag of adaptor1

Is that the best method to handle cutting both the adaptors?

Many thanks,
Phillipe

[chjiao]

A very good and valuable tool to cut adaptors from SOLiD sequencing data, but I wish that you could make the 5' adaptors cut for color-space data, since in the situation when Adaptors connected to another Adaptors this function is necessary.

[mmartin]

but I wish that you could make the 5' adaptors cut for color-space data, since in the situation when Adaptors connected to another Adaptors this function is necessary.

I have just released cutadapt 1.1, which adds this feature and is also 30% faster than before. See the release announcement at http://code.google.com/p/cutadapt/

[mmartin]

What's the best way to cutadapt Illumina reads where there are 2 adaptors such as:
5' - adaptor1 - sequence - adaptor2 - 3'

I want to the adaptors on both sides. What I have been doing is running cutadapt twice:
a. Run cut-adapt with -a flag with adaptor2
b. Feed the above output to another cut-adapt with -g flag of adaptor1

Is that the best method to handle cutting both the adaptors?

Yes, that is currently the best way to handle this kind of data. There's also an open feature request regarding some kind of 'paired adapter' mode, but this isn't implemented, yet.

[chjiao]

I have just released cutadapt 1.1, which adds this feature and is also 30% faster than before. See the release announcement at http://code.google.com/p/cutadapt/

Thanks, the new version of cutadap is challening when dealing with color-sapce data.

[kerhard]

cutadapt trimming length issue

Hi all,

I just used cutadapt to process some libraries I made and got a result that I don't quite understand. It appears that cutadapt trimmed more than 21 bases (the length of the adapter sequence) from many of the reads. I would think that the longest length that cutadapt could trim from the reads would be 21 bp if given a 21 bp-long adapter. Here is my output:

$ cutadapt -q 20 --quality-base=64 -a TCGTATGCCGTCTTCTGCTTG wt.fastq -o wt_adapttrim.fastq
cutadapt version 1.1
Command line parameters: -q 20 --quality-base=64 -a TCGTATGCCGTCTTCTGCTTG wt.fastq -o wt_adapttrim.fastq
Maximum error rate: 10.00%
Processed reads: 54310589
Trimmed reads: 8911163 ( 16.4%)
Total basepairs: 2111179389 (2111.2 Mbp)
Trimmed basepairs: 100971255 (101.0 Mbp) (4.78% of total)
Too short reads: 0 ( 0.0% of processed reads)
Too long reads: 0 ( 0.0% of processed reads)
Total time: 1918.18 s
Time per read: 0.04 ms

=== Adapter 1 ===

Adapter 'TCGTATGCCGTCTTCTGCTTG', length 21, was trimmed 8911163 times.

Lengths of removed sequences
length count expected
3 1088578 848603.0
4 735568 212150.7
5 672468 53037.7
6 645706 13259.4
7 570424 3314.9
8 507225 828.7
9 449549 207.2
10 387252 51.8
11 320108 12.9
12 298280 3.2
13 277658 0.8
14 277521 0.2
15 244603 0.1
16 239725 0.0
17 202783 0.0
18 205865 0.0
19 202092 0.0
20 233860 0.0
21 208748 0.0
22 136163 0.0
>=23 1006987 0.0


Any answers would be greatly appreciated!

This tool is super and very intuitive, thank you for it.

[kerhard]

Hi all,

I just used cutadapt to process some libraries I made and got a result that I don't quite understand. It appears that cutadapt trimmed more than 21 bases (the length of the adapter sequence) from many of the reads. I would think that the longest length that cutadapt could trim from the reads would be 21 bp if given a 21 bp-long adapter. Here is my output:

$ cutadapt -q 20 --quality-base=64 -a TCGTATGCCGTCTTCTGCTTG wt.fastq -o wt_adapttrim.fastq
cutadapt version 1.1
Command line parameters: -q 20 --quality-base=64 -a TCGTATGCCGTCTTCTGCTTG wt.fastq -o wt_adapttrim.fastq
Maximum error rate: 10.00%
Processed reads: 54310589
Trimmed reads: 8911163 ( 16.4%)
Total basepairs: 2111179389 (2111.2 Mbp)
Trimmed basepairs: 100971255 (101.0 Mbp) (4.78% of total)
Too short reads: 0 ( 0.0% of processed reads)
Too long reads: 0 ( 0.0% of processed reads)
Total time: 1918.18 s
Time per read: 0.04 ms

=== Adapter 1 ===

Adapter 'TCGTATGCCGTCTTCTGCTTG', length 21, was trimmed 8911163 times.

Lengths of removed sequences
length count expected
3 1088578 848603.0
4 735568 212150.7
5 672468 53037.7
6 645706 13259.4
7 570424 3314.9
8 507225 828.7
9 449549 207.2
10 387252 51.8
11 320108 12.9
12 298280 3.2
13 277658 0.8
14 277521 0.2
15 244603 0.1
16 239725 0.0
17 202783 0.0
18 205865 0.0
19 202092 0.0
20 233860 0.0
21 208748 0.0
22 136163 0.0
>=23 1006987 0.0


Any answers would be greatly appreciated!

This tool is super and very intuitive, thank you for it.

Sorry, I wasn't reading the description of the options carefully enough. I think I understand now what the above results indicate. Using the -a option and given an adapter 21 bp in length, a longer length than 21 bp would be trimmed by cutadapt from a given read if the adapter was found in the 5' end of the read, followed by more sequence.

If someone could verify if this is the case, or whether I am missing some other reason, it would be much appreciated.

[mmartin]

Sorry, I wasn't reading the description of the options carefully enough. I think I understand now what the above results indicate. Using the -a option and given an adapter 21 bp in length, a longer length than 21 bp would be trimmed by cutadapt from a given read if the adapter was found in the 5' end of the read, followed by more sequence.

If someone could verify if this is the case, or whether I am missing some other reason, it would be much appreciated.

Yes, that is correct. The column indicates the length of the removed sequence, which includes the bases after the adapter if there are any. It used to indicate the length of the matching adapter in earlier cutadapt versions, but I think that was less helpful.

[kerhard]

Yes, that is correct. The column indicates the length of the removed sequence, which includes the bases after the adapter if there are any. It used to indicate the length of the matching adapter in earlier cutadapt versions, but I think that was less helpful.

Thanks for confirmation. I'm glad I tried out cutadapt, as I was assuming my libraries were absent of adapter sequences, which turns out not to be true at all.

I suppose I still don't understand how some of these reads can have sequence AFTER the 3' adapters (eg., adapters found in the middle of the read). Searching for the full adapter sequence in the raw read files by hand, I notice that many times the sequences found after the 3' adapter are a string of A's. For example:

AGTCTADAPTERAAAAAAAAAAAA

TGCGTACGRACTADAPTERAAAAA

Any ideas as to what that means and how that may happen? Are these from the sequencing reactions on the Illumina machines or are these from library constructions?

[mmartin]

Any ideas as to what that means and how that may happen? Are these from the sequencing reactions on the Illumina machines or are these from library constructions?

I cannot really tell because I haven’t seen that before - but I mostly work with SOLiD data, so that doesn’t mean anything. You could look at the qualities that were assigned to the ”A“s. I would guess that they are from the library construction if the qualities are high, but that they are artifacts from the sequencing process if the qualities are low.

On further thought, I would also expect that artifacts from the sequencing process (the basecaller calling a base although there really is none) would lead to a sequence of random nucleotides, but I’m only speculating here.

[rimpi]

Can Cutadapt give the results in csfasta format.
I have solid data and I used cutadapt for removing the adapters in colorspace but there is an option of removing the adapters and then the resultant file in fastq format. can't I get it in csfasta format.
Here is the result of cutadaptIs it correct)
Command line parameters: -c -e 0.12 -a 330201030313112312 -x abc: --maq -o output.fastq /home/rimpi/solid/hubert/mergecs/1A.csfasta /home/rimpi/solid/hubert/mergecs/1A_QV.qual
Maximum error rate: 12.00%
Processed reads: 35052447
Trimmed reads: 23518246 ( 67.1%)
Total basepairs: 2593881078 (2593.9 Mbp)
Trimmed basepairs: 881159025 (881.2 Mbp) (33.97% of total)
Too short reads: 0 ( 0.0% of processed reads)
Too long reads: 0 ( 0.0% of processed reads)
Total time: 3386.48 s
Time per read: 0.10 ms

=== Adapter 1 ===

Adapter '330201030313112312', length 18, was trimmed 23518246 times.

Lengths of removed sequences
length count expected
3 294756 547694.5
4 122284 136923.6
5 155337 34230.9
6 196812 8557.7
7 260390 2139.4
8 318801 534.9
9 772603 133.7
10 419888 33.4
11 140579 8.4
12 214331 2.1
13 229820 0.5
14 456111 0.1
15 120289 0.0
16 107822 0.0
17 169000 0.0
18 235124 0.0
19 206157 0.0
>=20 19098142 0.0

[mmartin]

Sorry, csfasta/qual output is not supported at the moment. You will have to use a separate program to convert colorspace FASTQ to csfasta/qual (I think someone posted one in the forum). Don't use the --maq option if you do that. Which read mapper do you use that does not support colorspace FASTQ?

[minoru_harvest]

s silly question:
how can i pipeline cutadapt with other program ? cutadapt seems not being able to read from STDIN, so i cannot use "|"?