Question: sequencing depth and number of reads
0
gravatar for kanwarjag
4.5 years ago by
kanwarjag1.1k
United States
kanwarjag1.1k wrote:

i have an microRNA- seq data and I have applied a filter of >10 read counts in 90% of samples. I was wondering is it safe to say that 90 % of samples have sequencing depth of > 10 reads. Sequencing depth definition is little confusing Thanks

general • 1.5k views
ADD COMMENTlink modified 4.5 years ago by Brian Bushnell17k • written 4.5 years ago by kanwarjag1.1k

In which level of your analysis you are selecting ">10 read counts in 90%"? It looks to me that you are talking about filtering a gene expression matrix. Am i correct?

So The sequencing depth is another story. You can calculate the sequencing depth considering the reads aligned to the reference genome, the length of the reads of the sequencing machine (i.e. 101 for illumina) and the length of the genome of the organism you are using. Here the fromula:

original genome (G), the number of reads(N), and the average read length(L): (N*L)/G

But again, i do not think there is a link between sequencing depth and the filtering of the gene expression matrix (if that is what you are talking about).

ADD REPLYlink written 4.5 years ago by fusion.slope240

I am working with RPM as length of microRNA is very small

ADD REPLYlink written 4.5 years ago by kanwarjag1.1k

Sequencing depth is rather pointless in RNA-seq since the expression profile is totally not uniform, it's more useful to talk about the total number of reads that were sequenced.

ADD REPLYlink written 4.5 years ago by WouterDeCoster45k

if you want to check the splicing junction, and then possible splicing events like exon skipping or intron retention, sequencing depth is quite important.

ADD REPLYlink written 4.5 years ago by fusion.slope240
2

Yes, that's the local sequencing depth on that junction or for that exon and indeed that's important for discovery of the events you describe. But it's not "the sequencing depth" of the sample. Some genes will have far higher coverage than others so there is no uniform sequencing depth per samples. Just a total number of reads which is not uniformly distributed over the genome.

ADD REPLYlink written 4.5 years ago by WouterDeCoster45k
1
gravatar for Brian Bushnell
4.5 years ago by
Walnut Creek, USA
Brian Bushnell17k wrote:

I have applied a filter of >10 read counts in 90% of samples.

This statement seems to have no meaning. Can you clarify?

ADD COMMENTlink written 4.5 years ago by Brian Bushnell17k
1

From my interpretation this sentence is about the count table generated after performing feature counting, e.g. featurecounts, htseq-counts,... although the context in the first post is rather confusing. This filtering would say something about the read count per microRNA across and not about the samples an sich. If the entire sample would have <10 read counts that wouldn't really make sense. This looks like prefiltering (prior to differential expression analysis?).

ADD REPLYlink written 4.5 years ago by WouterDeCoster45k

exactly and that was my first reply but apparently the guy does not know in which part of the analysis he is..

ADD REPLYlink written 4.5 years ago by fusion.slope240

A base covered by at least 10 reads in each sample and this condition satisfied in at least 90% of the samples?

ADD REPLYlink written 4.5 years ago by GenoMax96k
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 1995 users visited in the last hour
_