Hi,
I currently did miRNA-mRNA predictions using RNA22 and RNAHybrid.
These programs report the minimum free energy (MFE) of the predicted interactions as a value judging the prediction. In general, RNAHybrid showed more interactions with lower MFE values.
Now my question: Is there a general MFE cutoff value that can be used to assess, if a predicted miRNA-mRNA binding is useful?
you could use the cut off of known miRNA targets such as let-7 which is around -27kcal/mol
Hi Markus
My experience is that DeltaG values could help in the ranking of the list, but there is not a limit or a threshold to consider, since this value depends on the GC content of the pairing region of the miRNA-mRNA hybrid, and on the presence of non-canonical binding sites. So, it is difficult to use just the DeltaG as a selection criterium.
I would advice you to use several target-prediction algorithms at the same time and select those targets which are predicted by a larger number of algorithms. You can use platforms as MirWalk (use the older 2.0 versio), which will allow you to select several target prediction engines simultaneously.
Hope it helps. Best of luck
Paco
The hybrid of miRNA:mRNA is considered perfect if the MFE value is ≤ -30 kcal/mol. However, useful binding could be greater than -30kcal/mol. Know that the threshold for MFE value is not fixed, some researchers used up to -15 kcal/mol. However, it seems to be high. I suggest an MFE value of ≤ -27 kcal/mol, similar to that of let-7 and CELF35-1 duplex. In addition to the MFE value, I would advise you to look at the binding pattern at the seed sequence of the miRNA (2–8 nucleotides from the 5’ end of the miRNA). If there is no gap in the alignment within the Watson and-Crick matching, then the binding of miRNA on its target mRNA is perfect.
Article In silico Analysis of Non-coding RNAs and Putative Target Ge...
- Badges
- Science topic
- Similar topics
- Filing