1. With Bismark

>>>/Users/bismark  --bowtie2 -N 1 -L 87 /User/hg_19 /Users/dataset-a.fastq (95.35%)

2. With bowtie2 independent

>>>bowtie2 -p 4 -x /Users/myIndex /Users/FILWET-2.fastq > output_best.sam (10.95%)

Hi,

Do you have paired end or single end data? For Bowtie2 you should designate paired data with -1 mate1.fq -2 mate2.fq OR for unpaired data -U single.fq

Did you set -N 1 when you ran Bowtie2 independent of bismark (default is N=0)? This is the number of mismatches allowed in the seed alignment. Option "L" in Bowtie2 is the length of the seed (default=22), which in the version I have has to be 3-32, so I'm not sure how you have -L 87 or if this is a bismark-specific option. Might be worth looking into.

Just out of curiosity, what kind of organims/sample is this from, and did you filter the data? 95% mapped is quite a high number.

Pathmanaban Ramasamy: Thank you so much for the insight. I think I'm incorrectly using the seed length. I also made sure it was bi-sulphite treated data. Both the reads and genome.

Cameron J Grisdale: They are single end reads, meaning i missed the -U option here. But here is my question, do I have to run bowtie2 in the with in its folder? or just a path is okay. I am trying to regenerate the results in the LAST paper https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3401460/ where he used hg19 and synthetic reads from http://cbrc3.cbrc.jp/~martin/dnemulator/bab/

My ultimate goal is to write a simple interactive python wrapper around bowtie2 to perform read alignment.

Any assistance will be greatly appreciated.

You don't need to run bowtie2 from within it's folder. You can either add the full path of the bowtie2 folder to your $PATH (usually in .bash_profile or .bashrc) or copy the bowtie2 executables into somewhere like /usr/bin/ or /usr/local/bin/ (sometimes this is done as part of the installation).

Have you tried running it with -U option to see if that's what caused issues?

What do you want the wrapper to do exactly?

>>>/Users/aamu_user/Desktop/bowtie2/bowtie2 -p 4 -x /Users/aamu_user/Desktop/hg_19/hg19 -N -L -U /Users/aamu_user/Desktop/hg_19/dataset-a.fastq -S output_file.sam

It is running as we speak. My task is to develop a pipeline for read alignment and detecting DMRs, genome wide.

The wrapper will help a user interact and set parameters.

1000000 reads; of these:

1000000 (100.00%) were unpaired; of these:

949360 (94.94%) aligned 0 times

46104 (4.61%) aligned exactly 1 time

4536 (0.45%) aligned >1 times

5.06% overall alignment rate

I'm not familiar with bi-sulphite data, but apparently wrappers are needed so that standard alignments software can deal with it. See: https://www.biostars.org/p/159903/ and this paper: http://bioinformatics.oxfordjournals.org/content/28/13/1698.long

Also, you should enter an integer number after the -N and -L options when running bowtie.