find all mate-pairs (75b / 0-infinity gap) alignment

[ramouz87]

Hi everyone,

I'm working with mate pair data (theoretical distance between pair is between 2 and 5 kb) and would be intrested to find all the mate pair that align independently of the Gap between pair.
I was thinking of aligning both reads separately using bowtie, the highest aligment among the 2 as a reference and then look at mate pair in the other alignment file.

Doing so, without optimization, i expect that running time would be less than one month(for 3 million reads) if we assume is that one search for mate pair takes 1 sec. i have a bi-xenon machine so i can optimize to divide time by 8. so it will takes 4 days. but i think that it's too long as bowtie do the alignement in few minute.

Any one have done this before ?
Any suggestion to solve this tqsk in decent time ?

Thanks in advance.
Regards,
Ramzi

[Jonathan]

Take a look at 'segemehl' -
http://seqanswers.com/wiki/Segemehl

Using Segemehl missmatches etc are quite cheap compared to bowtie/bwa/...
Disadvantage is the non-common mapping-out format

Best
-Jonathan

[ramouz87]

Take a look at 'segemehl' -
http://seqanswers.com/wiki/Segemehl

Using Segemehl missmatches etc are quite cheap compared to bowtie/bwa/...
Disadvantage is the non-common mapping-out format

Best
-Jonathan

Hi Jonathan,
Thanks for the suggestion,
i'm not sure but from the info in the link you provided the soft doesn't allow alignment to the full human genome at once but it would be nice if that application take in account paired-end data that will answer my question. but guess for now i have to write my own script my self.
Thanks again

Regards,
Ramzi

[nilshomer]

Hi everyone,

I'm working with mate pair data (theoretical distance between pair is between 2 and 5 kb) and would be intrested to find all the mate pair that align independently of the Gap between pair.
I was thinking of aligning both reads separately using bowtie, the highest aligment among the 2 as a reference and then look at mate pair in the other alignment file.

Doing so, without optimization, i expect that running time would be less than one month(for 3 million reads) if we assume is that one search for mate pair takes 1 sec. i have a bi-xenon machine so i can optimize to divide time by 8. so it will takes 4 days. but i think that it's too long as bowtie do the alignement in few minute.

Any one have done this before ?
Any suggestion to solve this tqsk in decent time ?

Thanks in advance.
Regards,
Ramzi

Try BFAST (I am the author). It will align the two independently although keeping the annotation consistent.

[ramouz87]

Try BFAST (I am the author). It will align the two independently although keeping the annotation consistent.

Hi Nils,
I have Bfast installed and i'm wondering how the parameters should be set to take in account all the mate-pair without a limit in the gap between them, the aim after that is to look for SVs.
Thanks in advance.
Ramzi

[nilshomer]

Hi Nils,
I have Bfast installed and i'm wondering how the parameters should be set to take in account all the mate-pair without a limit in the gap between them, the aim after that is to look for SVs.
Thanks in advance.
Ramzi

Use the recommended parameters found in the manual. The above is a property of the algorithm, not the settings. Let me know how it goes.