ChEMBL Resources


Friday, 29 May 2015

Compound popularity contest

Have you ever wondered which compound is the most popular in ChEMBL? And by popular I don't mean the one which cracks the best jokes at dinner parties; I mean the compound with the largest number of structural analogues or nearest neighbours (NNs). This number also gives an indication of the sparsity or density of the chemical space around a compound and is a useful concept during hit expansion and lead optimisation. 

This number of course depends on the fingerprint, the hashing and folding parameters, the similarity coefficient and the threshold. So let's say 2048-bit RDKit Morgan fingerprints with a radius of 2 or 3 (equivalent to ECFP_4 or ECFP_6) and Tanimoto threshold of 0.5. Why so low threshold? For an explanation, see here and here.

To calculate this compound 'popularity', one would need to calculate the full similarity matrix of the 1.4M compounds in ChEMBL. This used to be prohibitively computationally expensive just a few years ago; nowadays, thanks to chemfp, it only takes 5 commands and a few hours to calculate the matrix and counts at a certain similarity threshold on a decent machine. Here's how it's done:

#install chemfp
pip install chemfp

#get the chembl sdf

#calculate rdkit morgan fps
rdkit2fps --morgan --radius 3 --id-tag "chembl_id" --errors report chembl_20.sdf.gz -o rdkit_chembl.fps

#calculate counts of neighbours at a given threshold, here 0.5
simsearch --threshold 0.5 --NxN -c rdkit_chembl.fps -o chembl_sim_matrix.txt

#sort by number of neighbours and find most 'popular' compound

sort -rn chembl_sim_matrix.txt | head -25

Here's a list of the most popular ones with the number of NNs:

1570 CHEMBL3037891
1540 CHEMBL2373012
1523 CHEMBL3086861
1461 CHEMBL2158309
1451 CHEMBL440810
1451 CHEMBL414360
1428 CHEMBL2401865
1425 CHEMBL408133
1425 CHEMBL354100
1425 CHEMBL344931
1425 CHEMBL310737
1425 CHEMBL303256
1425 CHEMBL302290
1425 CHEMBL27867
1425 CHEMBL137783
1403 CHEMBL1866472
1396 CHEMBL3102922
1392 CHEMBL3102921
1392 CHEMBL3102920
1387 CHEMBL91573
1387 CHEMBL70362
1387 CHEMBL601767
1387 CHEMBL533732
1387 CHEMBL503883
1387 CHEMBL477

Most of the most popular compounds are large, usually peptides, e.g. CHEMBL3037891. We'll leave it as an exercise for the reader why that is. But this not always the case, for example CHEMBL477 is adenosine and it genuinely has a lot of near neighbours in ChEMBL (including a lot of stereo variants).  

And what about the least popular compounds, i.e. the ones with no neighbours above 0.5 Tanimoto? One would not expect a lot of them, since ChEMBL by design contains compounds from congeneric series as reported in the med. chem. literature. And yet, due to small size, symmetry (CHEMBL100050), possible mistakes, and peculiarities of fingerprint similarity, or due to being genuinely lonely (CHEMBL35416), they do exist.  

Apart from popularity contests, a similarity/distance matrix of ChEMBL compounds is the first step for clustering and graph/community detection analysis.


PS: Thebacon has only 22 neighbours...

Friday, 15 May 2015

What's Going On?

I’ve been asked a lot by mail recently ‘What’s Going On?’ Well, here is are some facts and some emotion.

So today is my last day at work here at EMBL-EBI. It’s been a fun and thrilling ride (for me at least), I’ve made lots of new friends, living life as an Open Data advocate and academic researcher, and most importantly having the privilege to lead the team here responsible for the ChEMBL database. It had been a long-term goal of mine to unlock large-scale bioactivity data from proprietary data silos and eye-wateringly expensive paywalls; so as US President George Bush famously said ‘Mission Accomplished!’. The impact of ChEMBL on academia, SMEs and large pharma has been great - and you can see the impact in new method development, but more importantly in potential new future drugs. My personal indebtedness to the Wellcome Trust for their support is immeasurable. An additional big shout out to Digital Science for their vision in donating the SureChEMBL platform to us.

I’ll be starting a new blog, covering my next adventure - artificial intelligence-enhanced drug discovery. I’ll tweet when this is up and running, but the first few weeks at a new job, as I’m sure you know, is spent sorting out pencils, working out where the best coffee is hidden and most importantly navigating the office politics of the milk in the fridge. When this blog starts up, I’ll tweet the url. For those of you interested in the ChEMBL groups activities, make sure you follow @ChEMBL and @SureChEMBL. If you want to see what I'm up to next, it's at @StratMed.

If any of you are ever in the West End of London (which to non-native Londoners actually means the centre) get in touch with me, and I’ll try to treat you to an orange mocha frappuccino.

Now for the bit you all actually care about….

  • Anne Hersey is taking over the ChEMBL Wellcome Trust Strategic Award grant for ChEMBL (which also covers SureChEMBL). Many of you will know Anne already, and know just what good news this is. Anne is also taking over the majority of our other grants and activities of the group, including our participation in IMI eTox, NIH IDG KMC, & GSK CTTV grants.
  • Jo McEntyre will be PI for EMBL-EBI on the IMI OpenPHACTS grant, although the majority of the work will be done by ChEMBL group staff. If you don’t know about the Open PHACTS platform, check out what they have done!
  • Ugis Sarkans will become PI for EMBL-EBI on the FP7 HeCaTos grant. The data content and modelling components will be done by ChEMBL group staff.

If you have general questions about ChEMBL or SureChEMBL first try the support email addresses and