new gene discovery

Question

[songmm19900210]

I want to blast my circRNAs which are identified by CIRI software to circRNAs which in circBase database(http://www.circbase.org/). how can i find new circRNAs from my identified circRNAs which are not in circBase database?
thanks!

Answer 1

[songmm19900210]

thanks Amitm, after blast with the fasta of the Circbase data, how to discoevey new circRNA? What is the criterion?

[Amitm]

hi,

I'm sorry I do not have any experience in interpreting circRNAs. The identification using sequences was a generic step hence replied initially. Having said that there are a few low hanging fruits you could go for if you haven't already -

1. Use BLAST E-value to select high-confidence candidates. (if you have too many to strt with)
2. Assuming you are working with human/ mouse, ENCODE is a wonderful resource to test hypothesis. For your high-confidence candidates, go to the UCSC browser, turn on ENCODE tracks and see if you have sign of promoter upstream (like histone marks, TFBS etc.).
3. Also you could check if the region has transcription signal in other ENCODE datasets as well

Answer 2

[Amitm]

hi,

Depending on the number of candidates you have, you could either use BLAST or bowtie. First you would need the fasta of the Circbase data, convert to a blast db or a bowtie index and then proceed.

If instead of fasta file, BED is available then use Galaxy: Extract Genomic DNA to get your db sequences first.