Question

Getting average fragment length for single end reads

2

Entering edit mode

7.5 years ago

Sreeraj Thamban ▴ 300

Hi, I have a set of raw files of RNASeq single end reads. Is there any tool for getting the average fragment length and SD from these files? I don't have access to a Bioanalyzer. I wanted to run Kallisto for these files. Thank you

RNA-Seq kallisto differential expresion • 9.3k views

ADD COMMENT • link updated 6.5 years ago by dark.lord ▴ 30 • written 7.5 years ago by Sreeraj Thamban ▴ 300

2

Entering edit mode

Prinseq and biopieces mainly

ADD REPLY • link 7.5 years ago by Buffo ★ 2.4k

1

Entering edit mode

Hello,

If I am not mistaken, these two pieces of software will give you statistics relating to reads and not fragments?

Which although useful, it is not what is requested by kallisto

ADD REPLY • link 7.1 years ago by Thomas ▴ 160

0

Entering edit mode

Thank you , i will try these two

ADD REPLY • link 7.5 years ago by Sreeraj Thamban ▴ 300

0

Entering edit mode

Is it wrong to use the one-liner provided here to calculate average fragment size and sd? Question: Mean and SD read length from a range of fastq files

ADD REPLY • link 6.5 years ago by dark.lord ▴ 30

0

Entering edit mode

There is no way to calculate average fragment size based on single-end sequencing data. You may be able to get a rough estimate based on Bioanalyzer/Tapestation results for the library that was sequenced.

ADD REPLY • link 6.5 years ago by GenoMax 147k

0

Entering edit mode

Is this not what MACS2 accomplishes using its sliding window approach?

ADD REPLY • link 4.4 years ago by millerh1 ▴ 40

score 4 · Answer 1 · 2017-05-22

4

Entering edit mode

7.5 years ago

James Ashmore ★ 3.5k

If you have single-end reads and don't have the Bioanalyzer output there is no way to calculate the actual fragment length. Instead just use a predicted value typical of most sequencing machines, say ~200bp?

ADD COMMENT • link 7.5 years ago by James Ashmore ★ 3.5k

1

Entering edit mode

Hi I used 150-200bp (best guess) as average fragment length in Kallisto but when I tried to reproduce the result from a tophat-cufflinks-cuffdiff pipeline the results shows very significant variation, but for paired-end reads, it was very consistent. So I think giving accurate fragment length is vital in Kallisto.