Is Tophat 2 good?

Question

[anonymous5982]

Hello, I am analysing a paper which used genome guided assembly with RNA-Seq data to assemble the transcriptome of some RNA. They used Tophat2 v.2.1.1 to do the assembly with the --b2-very-sensitive option. I have seen work which says that tophat is not great but I am a novice and don't really understand why.

Could someone please explain to me the limitations of using Tophat2?

Answer 1

[ATpoint]

The tophat tools are not assemblers but aligners, and both Tophat1 and 2 are deprecated (for years now). You should be using something recent, like STAR or hisat2 unless a legacy burden forces you not to.

[anonymous5982]

Oh okay thank you for clarifying! Is there a big reason why STAR or hisat2 outperforms Tophat?

[dsull]

The reason not to use it is performance. "TopHat2 is consistently among the worst performers on both human and malaria T2 and T3 libraries." https://t.co/deW5dgYc29 There are many other published benchmarks corroborating this result.

— Lior Pachter (@lpachter) April 24, 2019

(from one of the authors of the original TopHat; and read the paper linked to in that tweet).

[anonymous5982]

Thank you so much that is really helpful!

[Ram]

Hi, why did you delete the post?

[ATpoint]

I would start by reading the papers to learn about performance comparisons.