I need make all possible isomers and tautomers for a chemical library (about 10 000 compounds). But, I have never done such work. That is why, I don't know which software package is better to use.
1. Use the Markush Enumeration strategy by ChemAxon which is available in the program Marvin. You input a molecular structure with various R groups and variations and it will generate the structures. Further, for a single structure, you can generate all stereoisomers/tautomers in Marvin. This sometimes generates indiscriminately, so convert to InChI with OpenBabel (free) to remove duplicates helps.
A video on how to setup appropriate Markush input in Marvin can be found here:
https://www.youtube.com/watch?v=k8xDe1P_dgQ
2. Use EnuMol by the Akutsu laboratory. It is free online and you can download all structures as SMILES. You input a chemical formula with C, H, N, O and benzene rings. Use OpenBabel to convert SMILES to any other format. Limited in its use of hybridization on certain atoms but still very useful.
I've gotten pretty far with these tools! They all have their idiosyncratic limits, but you can obtain great starting results with them.
Hi Vasyl, take a look at MoKa which does this (protomers as well). It's commercial, so not free, but it's only a few hundred Euros for academics and is well respected. I work for the company that makes it - Molecular Discovery Ltd (www.moldiscovery.com), and there are several papers on the website that you can look at. Other well respected software includes ACD and ChemaAxon, but I don't think they are free. With MoKa you can also produce the most likely tautomers/protomers only, those within a range, or all of them. So depending on your work you can cut down the time a bit. If you need any more help don't hesitate to contact me ([email protected])
there is no software that will create ALL species. And the problem is differently complex based on which modification you are interested in.
Stereoisomers: easy to generate all, because there are 2 options per center, makes N^2. tripos stereoplex, schrodinger ligprep, or some of the other softwares already mentioned.
Double bond isomers: I am not aware of a software that does it.
Regioisomers: much more complex, e.g. placement of a methyl group ino,m,p position of a ring, or the product of a click reaction or of Diels Alder reaction etc. I guess you will have to create them amnually.
Tautomers: all algortihms I am aware of are rule-based with pattern matching. How elaborate a softwRe is, the patterns will always be incomplete, usually one substitution may yield the next matching pattern, the software will have to prune at some depth, find identical structures etc. And in the end, there is no easy way to evaluate the relative energies of the tautomers, and they then differ depending on the medium like gas, water, dmso etc.
Softwares are eg Tripos protoplex, schrodinger ligprep, pipeline pilot etc.
You may use www.chemicalize.org by ChemAxxon as an online tool. After structure input by smiles use "Properties Viewer". Here you can display stereoisomers, tautomers and more useful information.
I recommend LigPrep module of Schrodinger software, which can actually do what you are expecting for. Besides to create all the tautomers and protomers and isomers for a given molecule at desired pH, it can additionally optimize the resulting structure with the OPLS force-field. This is something I used long time ago, and I recall is the software originally used to generate the ZINC docking database for every purchasable commercial compound.
Unfortunately, it's not free nor even academics, but the worthy must be paid
If you are starting from an existing library of compounds, are you just looking for stereo-isomers, or conformers of these compounds, or are you wanting all structural isomers?
If you you are only interested in stereo-isomers of conformers, then starting from a SMILES string, it is fairly easy to generate them. Time and space constraints require that you might need to limit your expectations.
With this sort of requirement, recent versions of INTERCHEM, ahev this facility built in. With chirality specified tn the CIP notation, it is possible to generate stereo isomers with up to 4 of 5 chiral centres, and then order them according to energy.
You can down the latest version of the software (for Windows systems) from:
ftp://www.interprobe.co.uk. (See the website http://www.interprobe.co.uk).
1. Use the Markush Enumeration strategy by ChemAxon which is available in the program Marvin. You input a molecular structure with various R groups and variations and it will generate the structures. Further, for a single structure, you can generate all stereoisomers/tautomers in Marvin. This sometimes generates indiscriminately, so convert to InChI with OpenBabel (free) to remove duplicates helps.
A video on how to setup appropriate Markush input in Marvin can be found here:
https://www.youtube.com/watch?v=k8xDe1P_dgQ
2. Use EnuMol by the Akutsu laboratory. It is free online and you can download all structures as SMILES. You input a chemical formula with C, H, N, O and benzene rings. Use OpenBabel to convert SMILES to any other format. Limited in its use of hybridization on certain atoms but still very useful.
I've gotten pretty far with these tools! They all have their idiosyncratic limits, but you can obtain great starting results with them.