I am planning to set up a RNAseq sequencing run with the dual purpose of generating a new reference transcriptome and getting some differential expression data from an insect (genome size about 300mb). I will be extracting whole-body RNA from 4 developmental stages and two sexes, and doing three biological replicates for a total of 24 libraries.
I will be using the Illumina NextSeq platform, and I am trying to decide which kit to use. I had decided to go for the 300 Mid Output kit to get longer reads (150-150 paired end reads), but our core manager told me that to optimize such run, he would have to size select the library for fragments over 300bp to avoid loosing sequencing to overlapping paired end reads. The alternative he suggested was using a 75 High Output kit, which gives you many more, shorter reads.
There are advantages to using the 75 kit, including many more reads per sample, which improves the power for differential expression analysis, but I worry that I would end up with a very fragmented assembly.
I might say that I do have a genome and two transcriptomes for other species of the same genus, so a fragmented assembly might not be that much of an issue.
Have you tried using such short reads to de novo assemble NextSeq reads from eukaryotic RNA?
Hi Eduardo,
I've recently done de novo transcriptome assembly using Illumina for an insect comparing two developmental stages, but with the idea to get a first reference transcriptome. I used Myseq platform and the kit V3 600 (2x300) in order to build a paired-end library. I think that this is the newest one and you can obtain a high quality and longer reads (despite that is the most expensive too). I must to use that one because I do not have any close species to use as a reference genome. I obtained a great library with around 187.000 contigs with an average length of 870 bp. but in your case you have it, isn´t it?
So probably you are more interested in get more reads than in the length of them and the Output is 25M reads? Additionally, if the differential expression that you are expecting is not so "obvious", you should improve the number of differential reads in order to get the lowest expressed ones, considering you have genomes of related species that will help you with the annotation.
Regards and good luck!!
It can be done, and we have done something similar in our GermlncRNA project. PM me.
In my master Degree I use it for assembly from mtDNA in Fungi and the results are very well !
Thanks for your replies. I will give it a try and report back. As soon as I get the data back I will contact you, Tin-Lap.
A late update: I ended up sequencing single end, 75bp, and assembling with both Trinity and Velvet to obtain a quite decent assembly. I have yet to evaluate how fragmented is the assembly (I actually did two different experiments on two species, one sequencing 300M reads, and the other 30M, so I have a way of comparing assembly metrics vs coverage depth). But at the end of the day, I can use both as sequence databases, and was able to find the genes I was looking for - indeed, the main purpose of these transcriptomes was to avoid doing any more degenerate PCR.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics