Pharmacology A 'How To' guide on drug discovery

[AlsoTapered]

1) Download ChemSketch
2) Draw putative ligand
3) Create SMILES definition (tools/generate/SMILES definition)
4) Copy SMILES definition
5) Paste SMILES definition into PubChem
6) If their isn't a precise match, look at SIMILAR and SUBSTRUCTURE

Remember than chemicals very often had chiral centres and that these 'mirror images' may have significantly different activity profiles.
That way we can discard 99% of compounds BEFORE posting. Believe me, I've even approached BL with source-code to allow the site to display chemicals in their 3D structure. But of course, I presume they use an OTC BBS and nobody is being paid to add such features. We cannot expect them to solve this.

7)Upload image with both IUPAC & SMILES definitions as text, not part of the image.

Sadly ChemSketch doesn't visualize lone-pairs and so that is the 1 thing you will need to learn. VSEPR theory cover it, so one you have that, you can then begin to building training-sets. Do not underestimate training-sets. I had to laugh when someone presented dozens of methylphenidate derivatives... because I know 1 BLer had spend 5 years studying just that.

Choose a scaffold and work on it.

 

[arrall]

I've even approached BL with source-code to allow the site to display chemicals in their 3D structure

That’s the first I’ve heard of this.
Please PM me about it.

 

[AlsoTapered]

There are posts wish stickies at the top of this page.

 

[SpiralusSancti]

Believe me, I've even approached BL with source-code to allow the site to display chemicals in their 3D structure. But of course, I presume they use an OTC BBS and nobody is being paid to add such features. We cannot expect them to solve this.

that would really help people like myself to grasp some things better

 

[AlsoTapered]

SMILES is very compact and their is a free javascript to SMILES<-->IUPAC the data and another (admittedly primitive) one that SMILES-->3D image. Now their is also free software to calculate minimum-energy conformations BUT it's a chunk of code written for UNIX based machines.

I mean, if the software dragged the SMILES into PubChem then the 3D image could be captured. The problem here is that the code for all of the above spans 6 different languages making it potentially messy to keep updated.

As I have said elsewhere, investing £200 on Chem Office is a bargain when you look at the cost of the new on-line versions. The copies on Ebay are legit (unused key to activate) and the new version's only 'advantage' is dragging the user to the article... to pay $30-$50 a pop.

 

[Delmonte421]

that would really help people like myself to grasp some things better

You should buy this and play around with it like legos. That should help you understand r/s chiralities and L vs D in amino acids and what the differences is. Also learning about what groups take priority over other when naming and looking at the chirality of a molecule, kinda like in physics the right hand rule to help identify the different forced and how they act on each other.

 

[Delmonte421]

SMILES is very compact and their is a free javascript to SMILES<-->IUPAC the data and another (admittedly primitive) one that SMILES-->3D image. Now their is also free software to calculate minimum-energy conformations BUT it's a chunk of code written for UNIX based machines.

I mean, if the software dragged the SMILES into PubChem then the 3D image could be captured. The problem here is that the code for all of the above spans 6 different languages making it potentially messy to keep updated.

As I have said elsewhere, investing £200 on Chem Office is a bargain when you look at the cost of the new on-line versions. The copies on Ebay are legit (unused key to activate) and the new version's only 'advantage' is dragging the user to the article... to pay $30-$50 a pop.

Does IUPAC change yearly or is it still the same set of rules as I learned 20 years ago?

 

[AlsoTapered]

I HOPE people recognise stereo-isomerism (not to say positional isomerism). It's conformational isomerism and the physical properties of the molecule that are more nuanced. But Lipinski's Rule of Five that is the simplest way to test in-silico if a compound will be orally active.

IUPAC is a large topic with committees and subcommittees constantly discussing revisions. No older form of IUPAC is less precise than any later form but it has undergone constant modification.

But who uses IUPAC these days? If you can read this, you can download software that will convert any IUPAC into SMILES or into an image. But while SMILES isn't EASY to read manually, it's possible to do.

 

[Delmonte421]

But Lipinski's Rule of Five that is the simplest way to test in-silico if a compound will be orally active.

Most people on here I would assume haven't taken intro to Orgo let alone 500 level med chem classes.

 

[AlsoTapered]

Well they can all read the Wikipedia page, can't they?

Free online 'druglikeness' calculators based on RO5 are everywhere, it seems.

Molsoft L.L.C.: Drug-Likeness and molecular property prediction.

 

[Delmonte421]

Lipinski's rule states that, in general, an orally active drug has no more than one violation of the following criteria:[9]

• No more than 5 hydrogen bond donors (the total number of nitrogen–hydrogen and oxygen–hydrogen bonds)
• No more than 10 hydrogen bond acceptors (all nitrogen or oxygen atoms)
• A molecular mass less than 500 daltons
• A calculated octanol-water partition coefficient (Clog P) that does not exceed 5

Direct from the Wikipedia page, if you have no science background, do you think someone would understand that? Were you not a TA in graduate school lol or maybe all your students were a lot more brilliant then mine lol

 

[AlsoTapered]

Each detail is covered in even the most basic textbook on medicinal chemistry and only the LogP won't already be known by anyone who has obtained an undergraduate degree in chemistry. I would expect a student to be able to read.

That's why the on-line thingy is so useful - it let's people test compounds known to be active and to let people see how a modification alters those values.

It's the conformation of non-rigid compounds that requires more thought. Active conformation may require a local minimum-energy state or even an energy-state slightly above minimum given that the body provides a certain amount of energy.

 

[Delmonte421]

I agree with you, but I guess I had some real bad experiences catching students coping each other labs and cheating off each other that I just figure the majority of students are just assholes that want the easiest way to get things done instead of doing it the right way. I then transact that to humans in general, always taking the path of least resistance

 

[AlsoTapered]

Oh - the solution to that is NOT setting a single task to more than one student. It requires a strategy, but I have saved myself countless thousands of hours or work by using tuition simply as a form of delegation.

All of those young, well scrubbed faces polishing the metaphorical apple.

It would appear to be a skill required to obtain the most holy of holies - tenure.

But given that it's been 28 years since I HAD a student, maybe times change? Educating those WITH tenure is the impossible task. If you find out how, TELL ME.

 

[sekio]

I recommend that instead of ChemSketch you go the whole hog and "acquire" a copy of PerkinElmer ChemOffice.

ChemDraw is, frankly speaking, wonderful for the lazy chemist. You can draw whatever structure, have it figure our (R/S) and cis/trans chirality automagically, figure out MW and even do property prediction (incl. LogP and cLogP), clean up a weird-looking drawing, and best of all it does both IUPAC to structure, and structure-to-IUPAC conversion with the click of a button, and even updates the name on the fly as you edit the compound.
You even get clipart of common lab ware to diagram apparatus quickly and easily.
Also it comes with Chem3D too, you can import your 2D drawings and run MM2/MMF94 minimization on them, to get energy minimized accurate 3d models. And you can have the 2D structure display alongside, which auto-updates as you edit, and also vice versa.

Basically, it's a Real Cool Tool. PM for links.

sekio

 

[4meSM]

I recommend that instead of ChemSketch you go the whole hog and "acquire" a copy of PerkinElmer ChemOffice.

ChemDraw is, frankly speaking, wonderful for the lazy chemist. You can draw whatever structure, have it figure our (R/S) and cis/trans chirality automagically, figure out MW and even do property prediction (incl. LogP and cLogP), clean up a weird-looking drawing, and best of all it does both IUPAC to structure, and structure-to-IUPAC conversion with the click of a button, and even updates the name on the fly as you edit the compound.
You even get clipart of common lab ware to diagram apparatus quickly and easily.
Also it comes with Chem3D too, you can import your 2D drawings and run MM2/MMF94 minimization on them, to get energy minimized accurate 3d models. And you can have the 2D structure display alongside, which auto-updates as you edit, and also vice versa.

Basically, it's a Real Cool Tool. PM for links.

sekio

That’s what I use as well, wayy better than chemsketch. Do you happen to have a version for macOS? I have it in my windows computer but recently bought a macbook and couldn’t find the corresponding version. So now I’m forced to switch to my windows pc everytime I need to use chemdraw and mestrenova

 

[AlsoTapered]

Chemoffice? You can buy legitimate versions of the earlier editions as now it's a download... but a version a few years old is still an amazing tool.

 

[Flynnal]

Does anyone here know about vinylbital or its sodium salt?
It would be interesting to hear what others' opinions are on this specific and obscure barbiturate.
I've yet to find any real information about this drug, but it is supposedly the most toxic oral barbiturate, even more toxic than Seconal and therefore definitely more toxic than Nembutal/Amytal, which says something.

 

[Flynnal]

Chemoffice? You can buy legitimate versions of the earlier editions as now it's a download... but a version a few years old is still an amazing tool.

I'm interested in looking into this. I've been trying to find more info on vinylbital but have come up short, the only thing I have is 3 part (2 landscape pages per part) document series by a Vincent Badeux (have no idea who he is, other than I had this on my disk a long time ago) on an intro to barbs and toxicity, a set of barb derivatives and relationships between a set of barb derivatives, and vinylbital is featured in all 3 of those parts. The thing about these documents is they are so well polished, but who would have thought the French were not creative and colourful in their graphic/textural works. The docs I have prove the point, and I can upload those since they are public domain (but ironically no where to be found). Lucky I kept the copies safe.

Aside from that, and a slew of different brand names (Bykonox is one, Speda is another, but Vincent Badeux only refers to Optanox and Speda, and outside of that series there is no real information about the drug where it is actually compared to other barbiturates, other than "similar to pentobarbital" or "similar abuse potential to pentobarbital or amobarbital" with little else to say about it.

Apparently 900mg from the 0.9g minimum lethal dose listed in the table of 7 barbiturates (including pento and seco), killed a woman "with a history of heavy drug misuse" back in the 1960s. So it has to be really toxic to do that, right?

 

[AlsoTapered]

ChemOffice won't point you to specific articles BUT one part of the suite, Chem3D will allow you to overlay multiple barbiturate drugs so that you can see how the 5,5-disubstitution patterns overlay.

Vinylbital is interesting in that it partially overlays phenobarbitone (the alkene) and partly overlays Seconal so I presume it's of reasonable potency and intermediate action.

The serious problem with barbiturates is that tolerance to it's subjective effects and dependence quickly kick in BUT unlike most drugs, the body doesn't develop tolerance to it's toxicity so the therapeutic window, small already, closes.

PubChem will prove all the reference you need,