Tweet
Hello All,
This is my first post, and I'm newish to assembly, so stick with me.
I was curious as to what people out there are doing with regards to evaluating de novo transcriptome assemblies. I have tried read-remapping to assembled ests, pairing percentages (assuming you started with paired reads, lets say illumina), assembled est mapping to best references (to see how well your set represents whats out there), simple chimera searching by blasting against uniprot and looking for disjoint reference hits, frameshifts, and strand switches, and using coding and non-coding predictors (estscan, Infernal, Augustus) to determine what percentage of your set looks to be an est.
Do these things make sense to anyone else?
What if you don't have a well represented reference to compare against?
What other things can I do?
Lets say that the starting set is illumina pe reads of a reasonable length (54-100).
Thank you for your time and commentary
This is my first post, and I'm newish to assembly, so stick with me.
I was curious as to what people out there are doing with regards to evaluating de novo transcriptome assemblies. I have tried read-remapping to assembled ests, pairing percentages (assuming you started with paired reads, lets say illumina), assembled est mapping to best references (to see how well your set represents whats out there), simple chimera searching by blasting against uniprot and looking for disjoint reference hits, frameshifts, and strand switches, and using coding and non-coding predictors (estscan, Infernal, Augustus) to determine what percentage of your set looks to be an est.
Do these things make sense to anyone else?
What if you don't have a well represented reference to compare against?
What other things can I do?
Lets say that the starting set is illumina pe reads of a reasonable length (54-100).
Thank you for your time and commentary