normalization bam file

Schelarina · 08-27-2014, 06:19 AM

Hello,
I have 3 smallRNA seq libraries mapped to the genome. I realized they are different in size. How do I normalize the libraries starting from bam files?
thanks

Brian Bushnell · 08-27-2014, 02:30 PM

Are you sure they need to be the same size? FPKM, for example, is already normalized by read count.

But if you DO want to make them the same size, you can subsample with Reformat to make them all the same size. I suggest subsampling from the original fastq, then mapping, rather than subsampling from a bam file.

dpryan · 08-28-2014, 12:57 AM

This is generally taken care of when you calculate differential expression. Any program meant to handle this (DESeq2, edgeR, etc.) computes a size factor that's used to correct for this. Have a look at something like miRdeep2 if you're new to looking at small RNAs.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
How to make .bam file to .bai file format under MS Window7 operating system?	Min Jung Lee	Bioinformatics	6	11-21-2014 01:06 PM
Convert merged BAM back to per lane BAM or FASTQ file	danielsbrewer	Bioinformatics	6	10-03-2013 08:29 AM
samtools sort bam file error: generate non-existent file	mediator	Bioinformatics	0	03-05-2012 09:42 PM
what is the file size for a 30X human genome sequencing file, raw and BAM?	RNA-seq	Illumina/Solexa	2	04-15-2011 12:27 PM

08-27-2014, 06:19 AM	#1
Schelarina Member Location: P Join Date: Apr 2014 Posts: 18	normalization bam file Hello, I have 3 smallRNA seq libraries mapped to the genome. I realized they are different in size. How do I normalize the libraries starting from bam files? thanks

08-27-2014, 02:30 PM	#2
Brian Bushnell Super Moderator Location: Walnut Creek, CA Join Date: Jan 2014 Posts: 2,707	Are you sure they need to be the same size? FPKM, for example, is already normalized by read count. But if you DO want to make them the same size, you can subsample with Reformat to make them all the same size. I suggest subsampling from the original fastq, then mapping, rather than subsampling from a bam file.

08-28-2014, 12:57 AM	#3
dpryan Devon Ryan Location: Freiburg, Germany Join Date: Jul 2011 Posts: 3,480	This is generally taken care of when you calculate differential expression. Any program meant to handle this (DESeq2, edgeR, etc.) computes a size factor that's used to correct for this. Have a look at something like miRdeep2 if you're new to looking at small RNAs.