Which type of microRNAs(Mature, precursor,) are sequenced by Illumina HiSeq and GA ?

[watermark]

Hi all,


I'm using TCGA dataset to analysis miRNAs expression in diffrent cancer. They used Illumina HiSeq and Illumina GA to sequence miRNA and then quantify the expression level. but when I look at the IDs of miRNA in level3 folder of TCGA, some of them seems to belong to miRNA precursor or stem loop (I checked via miRBase).
But now question for me is that, with Illumina technology, they sequnece all miRNAs or they just seprate mature miRNA from electrophoresis gel and then sequence it ?
I looked at the META data file for miRNA by TCGA. they describe like this : "

Ligation of linkers and reverse transcription of small RNAs "PCR with sequencing primers, size fractionation" Sequencing on Illumina GAIIx Alignment of reads to reference genome Read counts per mirna isoform Normalized expression per mirna gene

[TiborNagy]

In our lab we sequence all miRNAs. The 24nt length fraction is the most abundant.

[watermark]

Do you know about the TGCA miRNASeq data? they just sequenced mature miRNA or all miRNA ?
I couldn't find relative information about it.

[GenoMax]

Your best bet is to contact the submitting lab. The email addresses in the IDF file example on this page may be the ones to try: https://wiki.nci.nih.gov/display/TCGA/miRNASeq

[peterawe]

Both sequencers sequence whatsever in your library, meaning pool of DNA prepared in the lab that you put into the machine. TCGA, like others interested in smRNA/miRNA, prepare libraries using some molecular biology that specifically amplify short RNA fragments from the RNA collected from the sample.

The resulting sequencing data contains millions of very short sequences (16-36bp). These sequences are aligned against the genome - and annotated according to the transcript matching the position in the genome and counted.

TCGA just counts the number of transcript aligning within each precursor. They could have counted the numbers within the mature transcript but didn't. So; your data matrix contains the sum of reads aligning within in the precursor - most of these are from one arm (5p or 3p), but from which is impossible to tell from the data you have at hand. Like mentioned in a previous post, the isoform expression matrix also available from TCGA gives you arm-level resolution and is probably what you want. Hope this helps.