• Bluelight HOT THREADS
  • Let's Welcome Our NEW MEMBERS!

⫸STICKY⫷ SMILES <-> Image <-> IUPAC

clubcard

Bluelighter
Joined
Apr 12, 2013
Messages
1,401
I am interested to know if this free chemistry script (Java) or other free scripts along the same lines could be added to the Bluelight site? https://opsin.ch.cam.ac.uk

I am not to keen on the use of 'Druglikeness' calculators. I think presenting the results of Lipinski's Rule of 5 alongside the data produced by the above script would be of more value. http://www.niper.gov.in/pi_dev_tools/DruLiToWeb/DruLiTo_index.html

There is even a free version of CHARMM that runs under Linux. That would be a BIG task but the results are just fantastic. https://www.charmm.org/charmm/program/versions/

Even though the site is in Chinese, https://www.drugfuture.com/chemdata/<name> returns the original patents & index papers.

The advantage I see is that people could search for similar chemicals and once a chemical has been added to the system, everyone can see it.

I am prepared to help convert the QSAR data of the ACTIVE drugs in 'Opiates' by R.Lenz et al. It's amazingly thorough but more than 90% of the drugs are inactive. Some drugs not covered by 'Opiates' is covered by 'Opioid Analgesics - Chemistry and Receptors' by Casy and Parfitt. Even then, both books end in the 1980s so there are a lot of much newer compounds.
 

sekio

Moderator: N&PD
Staff member
Joined
Sep 14, 2009
Messages
21,204
Location
streets of simcity
I'm going to sticky this as it is a very useful resource for the aspiring molecule-doodler.
 

clubcard

Bluelighter
Joined
Apr 12, 2013
Messages
1,401


External Packages
CHARMM directly interfaces with the following packages:
These packages have to be obtained and installed separately.

Now CHARMM was written in 'modern' Fortran 96, a programming language which I know not a single thing. Has anyone had the thrill-seeking job of rewriting Fortran 96 into Java?
 

polymath

Bluelight Crew
Joined
Nov 4, 2010
Messages
1,786
Location
Northern Europe
Now CHARMM was written in 'modern' Fortran 96, a programming language which I know not a single thing. Has anyone had the thrill-seeking job of rewriting Fortran 96 into Java?
There are some programs, findable with Google search, that translate Fortran code to C++ or Java, but I find it hard to believe that they always always produce a functioning result without you doing some corrections by hand afterwards. I think Fortran makes it a bit easier to use parallel processing in scientific computation than C++. At least that was the impression I got when listening to a Fortran parallel computing lecture some years ago.
 

clubcard

Bluelighter
Joined
Apr 12, 2013
Messages
1,401
Nice one.

Now, some 8 years ago I talked to a couple of programmers who had written an IMAGE<->SMILES<->IUPAC converter in Java (I think). They were prepared to let any IP to use it for about £3000 which isn't cheap BUT did have a 5 year QC on it so any problems, they fixed. I reckon they were too nice to do well but I loved their work.

If WE stored everything as IUPAC and the board brought up images and allowed people to copy as SMILES, it would end us having to hose images and I HOPE it would allow people to exchange ideas more readily.
 

clubcard

Bluelighter
Joined
Apr 12, 2013
Messages
1,401
In full disclosure, I spent almost 20 years as a professional games programmer BUT I only ever wrote code in assembly language so I rarely wrote entire games. Others would profile their game and I would rewrite the most time consuming part of the code in optimized assembly language. Sadly, their is little work for people who can ONLY code in such a low-level language (although many of the games I optimized were pretty famous). I mentioned that someone had written a Java IUPAC<-->SMILES<-->Image and I believe that this site uses it:


Now, it IS possible to optimize Java by following a few simple rules. For the truly hardcore, it's possible to write Java AS bytecodes. For those interested, to exit Jazelle (or indeed Java of any flavour):

R0=R1=0
bytecode $ff


But the important thing is that the Java used by this site is free to all. I would LOVE to see an image of the molecule with the IUPAC name underneath it & SMILES below that. I suspect that SMILES itself can be compressed by the simple fact that elements do not appear equally so common elements use short bitstreams while uncommon elements can use longer bitstreams.

Sorry to go on about it but if the site also had an editor, it would allow people to record & produce SMILES & IUPAC naming all within the site. It would avoid images having to be stored on other sites. BL would be a complete resource in 1 place.
 

clubcard

Bluelighter
Joined
Apr 12, 2013
Messages
1,401
I HOPE I am trying to do the best for chemistry. Of course, I don't discuss anything I AM doing on-line but as the patents will show, pyeyzolam (the alcohol mimic) was & is mine. In fact, pynazolam is even more euphoric but I was concerned that it WAS so euphoric. I could see people getting into deep trouble really fast. I suppose a clandestine chemist is supposed to milk a great drug for all it's worth but I always did and have concerned myself with it's safety. Want to know what my criteria is?

'Would I be happy for my own son to take this'

Thus many things were thrown away. OK, my wife ALSO tried everything and even she thought pyeyzolam was just too damned goo.
 
Top