I have my plant small RNA data from illumina. Done all quality filtration steps and conserved miRNA predictions. Now I have to predict novel miRNAs. In order to align, reference genome is not available for my plant but only transcriptome. Please any one help me, should I proceed with the plants own reference transcirptome data or whole genome of related species for novel miRNA prediction?
Yes, you can use whole genome of related species for novel miRNA identification from small RNA-seq read data. Select closely related species genome and try miRDeep2 pipeline for novel miRNA finding.
Thank you Ganesh...
Can you please tell me, is there a huge variation in results if I use my plant's own transcriptome data or it wont be reliable result?
this is the more neck bottle if you doing smallRNA without genome especially to get the the precussor. As @Ganesh say you may use close genome with assume that the pre is same and the end of story you need to validate the result. talk of Novel you need to have mature, precussor (structure) and the MFE, MEFI of you study.
Hi Nadiya,
Dont use your transcriptomics data to predicte miRNAs. These mRNAs are often ORFs which does not make any sense while looking for novel miRNAs. Also don't use the related organism too. Every organism has different set of miRNAs (sRNAs) which will result in false predictions.
You can search in ncbi for ESTs of your plant and try to map them on those, look in RFam or PNRD for your species and map on those if you find some ncRNAs.
Use tools which does not use reference genome to predict miRNAs.
Once done further to validate your sRNAs so the target prediction too and try to validate using RNA-seq mIRNA expression correlations, degradome, literature search of PCR.
Don't use any of the method which you have mention above this will only result in predictions which will be rejected by the reviewers eventually.
Best
Hi Nadiya
As rightly said by Ashwani Jha try to get nucleotide sequences of your own reference genome like BAC clones and other shotgun genome sequences etc might be deposited in NCBI. Closely related genome cant be used for novel miRNA prediction which will create biases in miRNA classification and etc.
All the best
The best one is to have you own Genome, but sometime with limit of info and database of plant genome sure, we cannot do anything. most early paper was use the close ref because of this issue. I was identify in miRNA study have bias. Then i start look how to predect the posible miRNA by Genome and Degradome data. after that verfy at wet lab.
you should use a reference genome from the species you study. Otherwise you will not detect species spefic miRNAs or conserved miRNAs that differ slightly in sequence from other species. miRDeep depends on detecting the genome hairpin where smallRNA reads are aligned. If you use another species you just detect the miRNAs that match that species, not the novel ones in the species you study.
- Badges
- Science topic
- Similar topics
- Biological Science
- Biochemistry
- Biomolecules
- Nucleic Acids
- RNA
- Small RNA
- microRNA