Pathway analysis - speed of FGSEA versus Mitch

FGSEA is among the most used pathway enrichment tools due to its speed and straightforward design. While doing a comparison of different enrichment tools recently, I compared it against a tool called "mitch" and I noticed something interesting. Mitch was actually faster than fgsea. This is strange, because my own tests from 2020 published in BMC Bioinformatics showed that fgsea was >10x faster (fgsea version 1.11.1) (Fig 1).

Fig 1. Tests conducted in 2020 show FGSEA is 10x faster than mitch.

As emphasised in our 2020 paper, speed wasn't the main priority for our software. We were focused on enabling multi-dimensional analysis.

But given this curious observation, I did a systematic test of speed of fgsea and mitch using a real dataset to try and understand whether fgsea underwent some code changes that made it slower. The codes for my work are on GitHub, and I used the bioconductor docker images to repeat this across Bioconductor versions from 3-11 onwards. Before 3.10 is not available unfortunately.

What we see is that running processes in a docker container doesn't make them slower compared to native execution (Fig 2). Also mitch was faster than all fgsea tests done here. I am not sure about what changes fgsea had after fgsea v.1.11 but they have made a big difference to speed. Although fgsea has optimised C code for permutation based tests, that is computationally intensive. Mitch on the other had is written in pure R, but doesn't use permutations, it does an ANOVA-on-ranks test which is less computationally intensive but still slow because it uses base R functions.

Also notice how BioC3-11 fgsea is really slow. I repeated it separately and can confirm these numbers are correct.

Fig 2. Execution time for fgsea and mitch on TR1900X system.

From Fig 2 you can also see that parallelisation doesn't provide a linear speedup. Sure there are some parts of these tools that are single-threaded, but it looks like the overhead of managing >10 threads is high enough to cancel out the benefit of additional threads. Therefore, on this 16 thread Threadripper 1900X workstation, any more than 8 parallel threads is just a waste.

I also ran this test with BioC 3.19 on a newer 32 thread workstation (Threadripper PRO 5955WX), and the result was only slightly faster, in line the higher clockspeed and more PCIe lanes (Fig 3). 

Fig 3. fgsea and mitch on TR5995WX system (BioC3.19 docker image).


Based on this info in Fig 2 and Fig 3, the maximum number of parallel threads I would use for fgsea would be 8 and for mitch it would be 6. Using this smaller number of threads will also reduce RAM usage as compared to using all available threads.

Popular posts from this blog

Two subtle problems with over-representation analysis

Image
Over-representation analysis is a helpful and really frequently used technique for understanding trends from omics data and gene lists more generally. Just to demonstrate this, I tabulated the number of citations of some of the most popular web tools and packages, which reaches a massive 191k citations! But if you have used some of these tools, you will notice that they yield subtly different results. We were curiuous about that and ran a whole bunch of investigations into the internal workings of these tools, in particular clusterProfiler. We identified two subtle problems with clusterProfiler, which we unpack in detail in our new publication , but here I will give you a quick overview. Problem #1: The background problem The first one we call the “background problem,” because it involves the software eliminating large numbers of genes from the background list if they are not annotated as belonging to any category. This results in removing a large number of unannotated genes from the b...
Read more

Data analysis step 8: Pathway analysis with GSEA

Image
In our RNA-seq series so far we've performed differential analysis and generated some pretty graphs, showing thousands of differentially expressed genes after azacitidine treatment. In order to understand the biology underlying the differential gene expression profile, we need to perform pathway analysis. We use Gene Set Enrichment Analysis ( GSEA ) because it can detect pathway changes more sensitively and robustly than some methods. A 2013 paper compared a bunch of gene set analyses software with microarrays and is worth a look. Generate a rank file The rank file is a list of detected genes and a rank metric score. At the top of the list are genes with the "strongest" up-regulation, at the bottom of the list are the genes with the "strongest" down-regulation and the genes not changing are in the middle. The metric score I like to use is the sign of the fold change multiplied by the inverse of the p-value, although there may be better methods out there...
Read more

Uploading data to GEO - which method is faster?

If you have had to upload omics data to GEO before, you'll know it's a bit of a hassle and takes a long time. There are a few methods suggested by the GEO team if you are using the Unix command line: Using 'ncftp' ncftp set passive on set so-bufsize 33554432 open  ftp://geoftp:yourpasscode@ftp-private.ncbi.nlm.nih.gov cd uploads/your @mail.com_ yourfolder put -R Folder_with_submission_files Using 'lftp' lftp  ftp://geoftp:yourpasscode@ftp-private.ncbi.nlm.nih.gov cd uploads/ your @mail.com _ yourfolder mirror -R Folder_with_submission_files Using 'sftp' (expect slower transfer speeds since this method encrypts on-the-fly) sftp  geoftp @s ftp-private.ncbi.nlm.nih.gov password: yourpasscode cd uploads/ your @mail.com _ yourfolder mkdir new_geo_submission cd new_geo_submission put file_name Using 'ncftpput' (transfers from the command-line without entering an interactive shell) Usage example: ncftpput -F -R -z -u  geoftp  -p "yourpasscode...
Read more