Normalization: Genome-guided de novo assembly with STAR and trinity

Question

[lucask3]

Greetings I am new on working with genome-guided assembly.

I have a question regarding normalization. Trinity says it is required to do in silico normalization prior to run the genome-guided assembly when you have a large number of reads.

My question is does STAR do any sort of normalization during mapping or should I do the normalization prior to mapping with STAR? If I need to do it prior to mapping, how can I do it? Should I use the trinity insilico_read_normalization.pl ?

Thanks!

Answer 1

[h.mon]

STAR doesn't perform any normalization, you have to do the normalization with some other program. insilico_read_normalization.pl is fine for this task, you may also use BBnorm or ORNA, among others.

[lucask3]

Hi Another question!

I have 16 RNA-seq single read libraries where I have 8 for a condition and 8 for another.

Before mapping with STAR I want to run an in silico normalization using trinity. My question is can I run all my 16 libraries in one run of insilico_read_normalization.pl? Or should I run the program separately for each library?

Here is what I am doing for running all libraries at once:

$/TRINITY_HOME/util/insilico_read_normalization.pl \
--seqType fq --JM 100G --max_cov 50 --CPU 15 \
--output /home/a-m/username/trinity/norm_reads \
--SS_lib_type R \
--single trim_lib1.fq.gz,trim_lib2.fq.gz,trim_lib3.fq.gz,trim_lib4.fq.gz,trim_lib5.fq.gz,trim_lib6.fq.gz,trim_lib7.fq.gz,trim_lib8.fq.gz,trim_lib9.fq.gz,trim_lib10.fq.gz,trim_lib11.fq.gz,trim_lib12.fq.gz,trim_lib13.fq.gz,trim_lib14.fq.gz, trim_lib15.fq.gz,trim_lib16.fq.gz

When I run it I am getting only one fq file back. Is this correct?

Thank you!