miRNA sequencing question

Question

[delgrodel]

Hello! I have a quick question regarding TCGA's level 3 miRNA sequencing data.

I notice that the data lists unique count values for miRNAs that share identical mature sequences but are encoded from distinct genomic loci. As these miRNAs are derived from distinct precursors, I assume that the relative abundance of the precursor sequences is used to estimate the contribution each precursor makes to the mature sequence read count. For example, if the relative abundance of pre-mir-92a-1 to pre-mir-92a-2 is 2:1, then it can be surmised that approximately 2/3 of the mature miR-92a reads come from pre-mir-92a-1 and 1/3 from pre-mir-92a-2, assuming the precursors are of equal length.

For the TCGA data, does anyone know whether this was the method that was used to assign counts to unique miRNAs sharing the same mature sequence?

I apologize if the question appears naive. My knowledge in this area is quite limited.

Any help would be greatly appreciated!

Best,
Alan

Answer 1

[Lorena Pantano]

Hi Alan,

I think they used miRDeep2 and I think it just counts all reads in all places.

About assuming that the proportion of precursors it should be similar to the proportion of the mature ... well, it is difficult to say. It could be one mature/precursor is more stable than other, so it risky to assume the exact mechanisms. Normally in the precursor you will see small number of sequences. If the difference is big maybe is due to some isomiR that is specific for some of the precursor and that's the reason for the big difference in different precursors.

If you want to do it in a general way, it doesn't matter where the miRNA is generated because the function should be the same since it should be the same sequence. Of course, then you have the isomiRs that differ in the seed region for instance, and the function may change, but normally people just assume all isomiRs of the same miRNA do the same.

Hope this helps,

Cheers