Good day!
I have some predicted sequences that was not previously annotated in plant genome.
Than It was found that it's possible to BLAST these sequences using RNAseq data found in ncbi SRA archieve and to get the .sam file containing aligned results.
The question is: how to understand if a sequence is transcribed?
In the first case all the query sequence is covered by RNAseq reads, so I think it's definitelly YES.
But what about other cases?
What percentage of sequence must be covered to understand it's definitelly transcribed? Could there be gaps?
Igor V Gorbenko you can check for coding potential and verify with RNAseq reads for horizontal coverage & depth. But the ideal scenario to be 100 % sure is to design a primer from the predicted sequences and validate with QPCR.
You can never be 100% sure from RNA-Seq data if a gene/transcript is definitely transcribed in the cell. There are however, several methods to measure reads from RNA-Seq data and statistically “estimate” the likelihood of significance. There is a package called “DESeq2” in R, which uses a statistically robust method to perform such analyses. https://bioconductor.org/packages/release/bioc/html/DESeq2.html
A logical way would be to begin with statistical analysis of the reads and then validate some candidate genes/transcripts with pcr.
- Badges
- Science topic
- Similar topics
- Cognitive Science
- Thinking