Skip to main content

Generation : Overview

To sum up the previous post flood; generation of constitutional isomers from the elemental formula can be done by generating all partitions of the total 'free' valence of the heavy atoms. The overall scheme is shown here:

(click for bigger, as usual). So, for each formula, multiple partitions can be made, and each of these makes multiple sub-partitions, and each of these correspond to one or more molecules.

Now, I won't pretend that any of this is particularly novel. I am no doubt re-expressing the problem of generating all possible molecules in a slightly different way. Having tried (and failed) to implement published methods, this was the best I could come up with.

I suspect that there are many improvements that could be made to the algorithm, and the implementation of it. Getting something that works, even in a limited way, seems like progress, however :)


Anonymous said…
Hi Gilleain,
looking at the figures of the partitions it becomes clear that the deterministic generation of all possible isomers is an embarrassingly parallel problem.

Each partition can be handled as single problem and that means if you have 10000 partitions and 10000 CPUs (CUDA TESLA, SiCortex Supercomputer)
you could dedicate each problem to one CPU core or thread.

The problem with the old monolithic CDK deterministic isomer generator code was, that the (FORTRAN style) code can be easily parallelized, but the canonizer was extremely slow. So even having n-CPUs at hand would
not solve the speed problem.

But I think for the molecular space below 500 Da the fully parallelized version could solve most of the problems in a sufficient time frame (if above problem would be fixed and n-CPUs would be available).

gilleain said…
Hi Tobias,

You are right, it does look like it can be easily run in parallel.

One important thing, though, is that the number of partitions grows much more slowly than the number of structures - for the CnH2n series, the number of partitions is (42, 627, 5604) for C=(10, 20, 30). There are a LOT more C30H60 structures than 5604...

So, it might be that the natural 'unit' would be smaller - but the problem at the moment is that it is still checking within the set of children of each partition for isomorphism.

Still, it is a good idea.

Anonymous said…
"number of partitions is (42, 627, 5604) for C=(10, 20, 30). There are a LOT more C30H60 structures than 5604..."

....well you are right, lets say below 200-300 Da. There we go, the isomorphism tester is still the bottleneck, so a fast isomorphism tester version is still needed.

If you take a CUDA TESLA C1060 with 240 GPU like streaming processors and 80 GFlop/s double precision fp (or 1000 single precision floating point precision) it should be still faster than an 8 core (16 thread) Intel Core I7 which has around 40 GFLOP/s (double precision) and 80 GFLOPs (SP). The CUDA bottleneck can be the transfer from the CPU to the GPU.

In conclusion a massively parallelized code version distributing each partition to each core, using an ultrafast isomorphism tester, together with a versatile good-list and bad-list handler, bundled with with a proper NMR and MS and IR handler would be the way to go :-)

If I go to Wolfram Alpha and ask for the number of all isomers in the universe it still tells me: 42


Popular posts from this blog

How many isomers of C4H11N are there?

One of the most popular queries that lands people at this blog is about the isomers of C4H11N - which I suspect may be some kind of organic chemistry question on student homework. In any case, this post will describe how to find all members of a small space like this by hand rather than using software.

Firstly, lets connect all the hydrogens to the heavy atoms (C and N, in this case). For example:

Now eleven hydrogens can be distributed among these five heavy atoms in various ways. In fact this is the problem of partitioning a number into a list of other numbers which I've talked about before. These partitions and (possible) fragment lists are shown here:

One thing to notice is that all partitions have to have 5 parts - even if one of those parts is 0. That's not strictly a partition anymore, but never mind. The other important point is that some of the partitions lead to multiple fragment lists - [3, 3, 2, 2, 1] could have a CH+NH2 or an NH+CH2.

The final step is to connect u…

The Gale-Ryser Theorem

This is a small aside. While reading a paper by Grüner, Laue, and Meringer on generation by homomorphism they mentioned the Gale-Ryser (GR) theorem. As it turns out, this is a nice small theorem closely related to the better known Erdős-Gallai (EG).

So, GR says that given two partitions of an integer (p and q) there exists a (0, 1) matrixA iff p*dominatesq such that the row sum vector r(A) = p and the column sum vector c(A) = q.

As with most mathematics, that's quite terse and full of terminology like 'dominates' : but it's relatively simple. Here is an example:

The partitions p and q are at the top left, they both sum to 10. Next, p is transposed to get p* = [5, 4, 1] and this is compared to q at the bottom left. Since the sum at each point in the sequence is greater (or equal) for p* than q, the former dominates. One possible matrix is at the top left with the row sum vector to the right, and the column sum vector below.

Finally, the matrix can be interpreted as a bi…

Havel-Hakimi Algorithm for Generating Graphs from Degree Sequences

A degree sequence is an ordered list of degrees for the vertices of a graph. For example, here are some graphs and their degree sequences:

Clearly, each graph has only one degree sequence, but the reverse is not true - one degree sequence can correspond to many graphs. Finally, an ordered sequence of numbers (d1 >= d2 >= ... >= dn > 0) may not be the degree sequence of a graph - in other words, it is not graphical.

The Havel-Hakimi (HH) theorem gives us a way to test a degree sequence to see if it is graphical or not. As a side-effect, a graph is produced that realises the sequence. Note that it only produces one graph, not all of them. It proceeds by attaching the first vertex of highest degree to the next set of high-degree vertices. If there are none left to attach to, it has either used up all the sequence to produce a graph, or the sequence was not graphical.

The image above shows the HH algorithm at work on the sequence [3, 3, 2, 2, 1, 1]. Unfortunately, this produce…