adapter trimming and length/otogenics

[lwhitmore]

I have received some chip seq data from the company otogenics. They provide two fastq files one that appears to contain the adapter sequence and one that does not based on fastqc reports. Additionally in these fastqc reports the read length does not change with the trimming of the adapter. I am wondering how this is possible and maybe what program they may have used to trim the adapter. Any help would be greatly appreciated.

Thanks
Leanne

[Brian Bushnell]

You can't trim adapters sequence without changing the read length, but you can either throw away reads that have adapter sequence, or (theoretically) produce reads that never had adapter sequence in the first place. Also, if you have for example a fragment library and a long mate pair library, they may have different adapters.

[paa6]

I have received some chip seq data from the company otogenics. They provide two fastq files one that appears to contain the adapter sequence and one that does not based on fastqc reports. Additionally in these fastqc reports the read length does not change with the trimming of the adapter. I am wondering how this is possible and maybe what program they may have used to trim the adapter. Any help would be greatly appreciated.

Thanks
Leanne

I have used fastx_toolkit for trimming adapters. http://hannonlab.cshl.edu/fastx_toolkit/
it's actually good and for trimming adapters u should use FASTA/Q Clipper.

[mastal]

What is the length of your reads, and do all the reads in each file have the same length according to the fastqc reports?

[lwhitmore]

mastal,
the length of my reads are 100b and all the reads have the same length in both the fastqc reports before and after trimming

[mastal]

What do the 2 fastq files represent, before and after trimming, or 2 different samples, or R1 and R2 of paired-end reads??

[lwhitmore]

before and after trimming on 1 sample for single end reads

[mastal]

How many reads in each sample?

[lwhitmore]

7811028 reads

[mastal]

Could you post the first few lines of each file?

It seems impossible that the reads would all have the same length after trimming as before if anything was actually trimmed, and if nothing was trimmed, you would expect the fastqc report to still show the adapter sequence in the over-represented reads.

[lwhitmore]

from the first file with the adaptor
@HWI-ST1129:515:H8V3LADXX:1:1101:2601:1960 1:N:0:CCGTCC
NAGAAATTTGGAAAATCAAATGCTTGAAGTAAGAGGACGATATTAAAACTTTTGTAACCAGAGACTACTTTAAGAAAAATCTGCTACTACTTTAACAAAG
+
#1DFFFHHHHHJJJJJJJJJJJJJJIJHHIIIEIHIHHGHIJJIJJJJJJJJIIJGJJJJJJJJJJJHHHHHHHFFFFEEDEEEEDDDDDEDDDDDDD
@HWI-ST1129:515:H8V3LADXX:1:1101:4187:1965 1:N:0:CCGTCC
NACAGAGCCTCGCTCTGTCTCCCAGGCTGGATGGAGTGCAGTGGCGCGATGTTGGCTCACTTCAAGCTCCGCGTCCTGTGTTCATGCCATTCTTCTGCCT
+
#1=DFFFFHHHHHJJJJJJJJJJJJJJJJJHIJJJJHIJIJGHJJJJJJJJJJJHHHHHFFFFFFEEEEEDDDDDDDDCDDDDEEDDDDDDEEEDDDDDD
@HWI-ST1129:515:H8V3LADXX:1:1101:4380:1977 1:N:0:CCGTCC
NTCCTCCCAAGAGAGATAGAGGAAGGAAAGGGAGAGATGGGACCACCACAGTGAGCAAATGGATCAGATTATTACTCTAAAATGTTCTTTTAGATCGGAA


From the second fastq file without the adaptor
ACAATGACACTTAGCATTTACTGTGTTAGTTAACATTTAGCAGATCTTTGTTAAAGTAGTAGCAGATTTTTCTTAAAGTAGTCTCTGGTTACAAAAGTTT
+
CC@FFFFFHHHHHJJJJJJIJJIFHIIJJHIIJJJJIJJJIJIJJJJJIJIJJJJJBGIIGIJJJJJIJJJJIJJJJJCHIEIIIIJH?HEHHFDFFCEE
@HWI-ST1129:515:H8V3LADXX:1:1101:4187:1965 2:N:0:CCGTCC
ATTAGCCGGGCATAGTGGCAGGGGCCTGTAGTCCCAGCTACTCGGTAGGCTGAGGCAGAAGAATGGCATGAACACAGGGCGCGGGGCTTGAAGTGAGCCA
+
@@@DDDDDHHHHHIIHIIIIII0??D1)9990?DH#################################################################
@HWI-ST1129:515:H8V3LADXX:1:1101:4380:1977 2:N:0:CCGTCC
AAAAGAACATTTTAGAGTAATAATCTGATCCATTTGCTCACTGTGGTGGTCCCATCTCTCCCTTTCCTTCCTCTATCTCTCTTGGGAGGAAAGATCGGAA

[mastal]

Looks like you have paired-reads from the same sample.

This part of the header - 2:N:0:CCGTCC - tells you it's the second read of a pair.

So if your reads are all the same length, that would suggest that they haven't been trimmed.

[lwhitmore]

ahh ok sorry that i didn't pick up on that i am very new to sequence analysis.

One more question if you don't mind
Why wouldn't the second file have an over represneted sequence (or an adapter)?

Thanks Again!

[mastal]

the sequences in the R2 file will be different from the sequences in the R1 file.

it's possible that there are other over-represented sequences that are more abundant in the R2 file, so the adapter sequence doesn't make it into the top over-represented sequences.

Is it the FastQC report that has flagged the sequences as adapter sequences?

[lwhitmore]

Yes it was the fastqc report that flagged the sequence as an adapter