# So much to do, so little time

Trying to squeeze sense out of chemical data

## vSDC, Rank Products and DUD-E

This post is a follow-up to my previous discussion on a paper by Chaput et al. The gist of that paper was that in a virtual screening scenario where a small number of hits are to be selected for followup, one could use an ensemble of docking methods, identify compounds whose scores were beyond 2SD of the mean for each method and take the intersection. My post suggested that a non-parametric approach (rank products, RP) performed similarly to the parametric approach of Chaput et al on the two targets they screened.

The authors also performed a benchmark comparison of their consensus method (vSDC) versus the individual docking methods for 102 DUD-E targets. I was able to obtain the individual docking scores (Glide, Surflex, FlexX and GOLD) for each of the targets, with the aim of applying the rank product method described previously.

In short, I reproduced Figure 6A (excluding the curve for vSDC). In
this figure, $$n_{test}$$ is the number of compounds selected (from the ranked list, either by individual docking scores or by the rank product) and $$T_{h>0}$$ is the percentage of targets for which the $$n_{test}$$ selected compounds included one or more actives. Code is available here, but you’ll need to get in touch with the authors for the DUD-E docking scores.

As shown alongside, the RP method (as expected) outperforms the individual docking methods. And visual comparison with the original figure suggests that it also outperforms vSDC, especially at lower values of $$n_{test}$$. While I wouldn’t regard the better performance of RP compared to vSDC as a huge jump, the absence of a threshold certainly works in its favor.

One could certainly explore ranking approaches in more depth. As suggested by Abhik Seal, Borda or Condorcet methods could be examined (though the small number of docking methods, a.k.a., voter, could be problematic).

UPDATE: After a clarification from Liliane Mouawad it turns out there was a mistake in the ranking of the Surflex docking scores. Correcting that bug fixes my reproduction of Figure 6A so that the curves for individual docking methods match the original. But more interestingly, the performance of RP is now clearly better than every individual method and the vSDC method as well, at all values of $$n_{test}$$

Written by Rajarshi Guha

February 13th, 2016 at 7:25 pm

## Hit Selection When You’re Strapped for Cash

with one comment

I came across a paper from Chaput et al that describes an approach to hit selection from a virtual screen (using docking), when follow-up resources are limited (a common scenario in many academic labs). Their approach is based on using multiple docking programs. As they (and others) have pointed out, there is a wide divergence between the rankings of compounds generated using different programs. Hence the motivation for a consensus approach, based on the estimating the standard deviation (SD) of scores generated by a given program and computing the intersection of compounds whose scores are greater than 2 standard deviations from the mean, in each program. Based on this rule, they selected relatively few compounds – just 14 to 22, depending on the target and confirmed at least one of them for each target. This represents less than 0.5% of their screening deck.

However, their method is parametric – you need to select a SD threshold. I was interested in seeing whether a non-parametric, ranking based approach would allow one to retrieve a subset that included the actives identified by the authors. The method is essentially the rank product method applied to the docking scores. That is, the compounds are ranked based on their docking scores and the “ensemble rank” for a compound is the product of its ranks according to each of the four programs. In contrast to the original definition, I used a sum log rank to avoid overflow issues. So the ensemble rank for the $$i$$’th compound is given by

$$R_i = \sum_{j=1}^{4} \log r_{ij}$$

where $$r_{ij}$$ is the rank of the $$i$$’th compound in the $$j$$’th docking program. Compounds are then selected based on their ensemble rank. Obviously this doesn’t give you a selection per se. Instead, this allows you to select as many compounds as you want or need. Importantly, it allows you to introduce external factors (cost, synthetic feasibility, ADME properties, etc.) as additional rankings that can be included in the ensemble rank.

Using the docking scores for Calcineurin and Histone Binding Protein (Hbp) provided by Liliane Mouawad (though all the data really should’ve been included in the paper) I applied this method using the code below

 12345678910 library(stringr) d <- read.table('http://cmib.curie.fr/sites/u759/files/document/score_vs_cn.txt',                 header=TRUE, comment='') names(d) <- c('molid', 'Surflex', 'Glide', 'Flexx', 'GOLD') d$GOLD <- -1*d$GOLD ## Since higher scores are better ranks <- apply(d[,-1], 2, rank) lranks <- rowSums(log(ranks)) tmp <- data.frame(molid=d[,1], ranks, lrp=rp) tmp <- tmp[order(tmp$lrp),] which(str_detect(tmp$molid, 'ACTIVE'))

and identified the single active for Hbp at ensemble rank 8 and the three actives for Calcineurin at ranks 3, 5 and 25. Of course, if you were selecting only the top 3 you would’ve missed the Calcineurin hit and only have gotten 1/3 of the HBP hits. However, as the authors nicely showed, manual inspection of the binding poses is crucial to making an informed selection. The ranking is just a starting point.

Update: Docking scores for Calcineurin and Hbp are now available

Written by Rajarshi Guha

February 5th, 2016 at 1:36 am

## Chemistry, Clouds, Collaboration (Part 1)

There’s been an interesting discussion sparked by Deepaks post, asking why there is a much smaller showing of chemists and chemistry applications in the cloud compared to other life science areas. This post led to a FriendFeed thread that raised a number of issues.

At a high level one can easily point out factors such as licensing costs for the tools to do chemistry in the cloud, lack of standards in data sets and formats and so on. As Joerg pointed out in the FF thread, IP issues and security are major factors. Even though I’m not a cloud expert, I have read and heard of various cases where financial companies are using clouds. Whether their applications involves sensitive data I don’t know, but it seems that this is one area that is addressable (if not already addressed). As a side note, I was interested in seeing that Lilly seems to be making a move towards an Amazon based cloud infrastructure.

But when I read Deepaks post, the question that occurred to me was: what is the compelling chemistry application that would really make use of the cloud?

While things like molecular dynamics are not going to run too well on a cloud set up, problems that are data parallel can make excellent use of such a set up. Given that, some immediate applications include docking, virtual screening and so on. There have been a number of papers talking about the use of Grids for docking, so one could easily consider docking in the cloud. Virtual screening (using docking, machine learning etc) would be another application.

But the problem I see facing these efforts is that they tend to be project specific. In contrast doing something like BLAST in the cloud is more standardized – you send in a sequence and compare it to the usual standard databases of sequences. On the other hand, each docking project is different, in terms of receptor (though there’s less variation) and ligand libraries. So on the chemistry side, the input is much larger and more variable.

Similarity searching is another example – one usually searches against a public database or a corporate collection. If these are not in the cloud, making use of the cloud is not very practical. Furthermore, how many different collections should be stored and accessed in the cloud?

Following on from this, one could ask, are chemistry datasets really that large? I’d say, no. But I qualify this statement by noting that many projects are quite specific – a single receptor of interest and some focused library. Even if that library is 2 or 3  million compounds, it’s still not very large. For example, while working on the Ugi project with Jean-Claude Bradley I had to dock 500,000 compounds. It took a few days to set up the conformers and then 1.5 days to do the docking, on 8 machines. With the conformers in hand, we can rapidly redock against other targets. But 8 machines is really small. Would I want to do this in the cloud? Sure, if it was set up for me. But I’d still have to transfer 80GB of data (though Amazon has this now). So the data is not big enough that I can’t handle it.

So this leads to the question: what is big enough to make use of the cloud?

What about really large structure databases? Say PubChem and ChemSpider? While Amazon has made progress in this direction by hosting PubChem, chemistry still faces the problem that PubChem is not the whole chemical universe. There will invariably be portions of chemical space that are not represented in a database. On the other hand a community oriented database like ChemSpider could take on this role – it already contains PubChem, so one could consider groups putting in their collections of interest (yes, IP is an issue but I can be hopeful!) and expanding the coverage of chemical space.

So to summarize, why isn’t there more chemistry in the cloud? Some possibilities include

• Chemistry projects tend to be specific, in the sense that there aren’t a whole lot of “standard” collections
• Large structure databases are not in the cloud and if they are, still do not cover the whole of chemical space
• Many chemistry problems are not large in terms of data size, compared to other life science applications
• Cheminformatics is a much smaller community than bioinformatics, though is applies mainly to non-corporate settings (where the reverse is likely true)

Though I haven’t explicitly talked about the tools – that certainly plays a factor. While there are a number of Open Source solutions to various cheminformatics problems, many people use commercial tools and will want to use them in the cloud. So one factor that will need to be addressed is the vendors coming on board and supporting cloud style setups.

Written by Rajarshi Guha

February 22nd, 2009 at 5:00 pm