Finding translocations

[Bakhtiyar Taghizada]

Hi! I have sequenced entire genome of S.cerevisiae with Illumina and have paired end data. Now I am supposed to locate any translocation present in the genome. First I was planning to search for the regions that have higher coverage. My translocated strain did not sporulate and it just divided by mitosis as diploids. Since translocation results in transferring one part of chromosome to another, when I obtain fragments from entire genome to sequence, the number of translocated sequence doesnt change so the coverage should not be high in that regions. Now I am planning to search for the reads that have mate pairs in different chromosomes. Please tell me if my ideas are reasonable.

[Brian Bushnell]

Illumina coverage is very nonuniform, particularly with amplified data (which yours may or may not be); also, features like can cause high coverage. I think you'll get too much noise to see any signal.

Looking for mate pairs on different chromosomes is better. Perhaps you could map the reads, filter for only reads that are improper pairs, and then look at the coverage for just those - they should be high in areas of translocations, and very low everywhere else.

[Bakhtiyar Taghizada]

Thanks for your reply! I am confused in this point, why should I expect high coverage in translocated regions? How can it be considered as amplified sequences if it was just moved from one chromosome to another?

[woodydon]

Thanks for your reply! I am confused in this point, why should I expect high coverage in translocated regions? How can it be considered as amplified sequences if it was just moved from one chromosome to another?

You could describe the average coverage and sequencing length of your data to help people think more precisely.

[Brian Bushnell]

Thanks for your reply! I am confused in this point, why should I expect high coverage in translocated regions?

Improperly paired reads are more likely to come from translocations. Thus, if you filter and look at only the improper pairs, coverage will be higher near translocations. The properly paired reads won't help you much.

How can it be considered as amplified sequences if it was just moved from one chromosome to another?

Amplification is unrelated; it's part of library construction.

[Bakhtiyar Taghizada]

You could describe the average coverage and sequencing length of your data to help people think more precisely.

Average coverage is around 40 and read length is around 100.