puzzle picture
1519: HIV Protease Inhibitor Small Molecule Design: Round 2
Status: Closed

Summary

Name: 1519: HIV Protease Inhibitor Small Molecule Design: Round 2
Status: Closed
Created: 05/07/2018
Points: 0
Expired: 05/16/2018 - 23:00
Difficulty: Intermediate
Description: Note: Due to player reports of persistent crashing, this puzzle is now worth zero points, so players may skip this puzzle without any effect on their rankings. However, if players would like to continue playing the puzzle, we will still analyze the puzzle solutions. And of course, we appreciate all of the detailed feedback about the puzzle!

The goal of this puzzle is to change the chemical structure of the ligand in the center of the protein to find better inhibitors of HIV protease. It's less about changing the protein to fit the ligand, and more about changing the ligand to fit the protein. Use the Ligand Design tool in the selection interface (or the "Ligand Design" mode in the original interface) to alter the chemical identity of the inhibitor.
See this previous blog post for more information on small molecule design, and Puzzle 1432 for the previous round. In contrast with the previous round, we're starting with a smaller starting molecule and are running without a similarity filter.
Categories:

Top Groups

RankGroupScorePoints
1Contenders13,9730
2Anthropic Dreams13,9160
3Go Science13,8940
4Void Crushers13,8170
5Beta Folders13,7340

Top Evolvers

Top Soloists



Need this puzzle? Log in to download.  

Comments

Joined: 02/08/2012
Groups: None
here are some great links to get us all started

https://www.umassmed.edu/schifferlab/research/pathways-to-resistance/hiv-protease/

This link gives 9 approved inhibitors. I suggest trying to build them. Although we should probably build our own, building the already approved ones should help with making our own.

https://www.medscape.org/viewarticle/506280

This link might five more detail on the kind of protease inhibitors (PIs) to look for.

Joined: 12/27/2012
Groups: Beta Folders
the medscape link requires a subscription...

Unfortunately, the second article requires a subscription to view.

The first article shows the three-letter codes for the protease inhibitors. The codes and the generic names are shown below. See protease inhibitor on the Foldit wiki for more.

The codes and generic names are:

  • APV (Amprenavir)
  • ATV (Atazanavir)
  • DRV (Darunavir)
  • FPV (Fosamprenavir)
  • IDV (Indinavir)
  • LTV (Lopinavir)
  • NFV (Nelfinavir)
  • RTV (Ritonavir)
  • SQV (Saquinavir)
  • TPV (Tipranavir)

Edit: added generic names and included FPV/Fosamprenavir as discussed below. See protease inhibitor on the Foldit wiki for more.

toshiue's picture
User offline. Last seen 9 hours 40 min ago. Offline
Joined: 01/31/2016
Groups: Go Science
10th protease inhibitor - FPV

Found a 10th protease inhibitor on an older list, FPV (Fosamprenavir) which is rapidly metabolized into Amprenavir, hence its probably deletion from the list, but they are not chemically identical (APV = C25H35N3O6S, FPV (PDB: 3nu3) = C25H36N3O9PS). Difference may be of significance to us? (fwiw).

Joined: 09/22/2017
Groups: Beta Folders
Fosamprenavir

Fosamprenavir is a pro drug for Amprenavir - where you see that phosphate group a hydroxyl takes it's place in the active drug

Joined: 12/27/2012
Groups: Beta Folders
Codes can be tough to find...

Some of the three-letter codes for these protease inhibitors are a little obscure. They don't always correspond to the ligand name in the PDB.

The new article protease inhibitor on the wiki lists the generic drug name and the PDB ligand identifier.

Joined: 12/27/2012
Groups: Beta Folders
Some results....

I managed to build all nine PIs (protease inhibitors), not including Fosamprenavir.

The molecular weights of my versions match what the PDB reports to within a small fraction. The weights are measured in grams per mole. An extra gram per mole most likely means an extra hydrogen somewhere. Converting a single bond to a double bond should decrease molecular weight by 1 g/mol.

Three PIs had a big molecular weight penalty: -1123.04 for ATV (Atazanavir), -1364.45 for RTV (Ritonavir), and -612.825 for SQV (Saquinavir).

There was cLogP penalty of -700 for TPV (Tipranavir).

My versions probably don't have the right stereochemistry. I was mainly focused on getting the right atoms connected to each other, so I'm sure some things that should have been on the left ended up on the right.

I didn't have a lot of luck getting any of these PIs to interact with the protein. A much smaller molecule that I threw together did a lot better.

spvincent's picture
User offline. Last seen 2 hours 52 min ago. Offline
Joined: 12/07/2007
Groups: Contenders
A couple of issues

This puzzle crashes all the time for me in main (Ubuntu 16.04). No stack info: trying to reproduce.

Also frequently when trying to add an O or N I get a penalty from the Ligand cLogP filter of something absurd like 1.84467e21. Could probably get a test case on this if its not an already known problem.

Joined: 09/22/2017
Groups: Beta Folders
Having issues

everytime I goto upload a file to scientists my game crashes shortly after uploading

Windows 10 64 bit

Joined: 09/22/2017
Groups: Beta Folders
error message

Intermittent crashes with the following in the logfile

1: alIsAuxiliaryEffectSlot +598376 bytes (no line)
2: alIsAuxiliaryEffectSlot +598984 bytes (no line)
3: library_main +5297766 bytes (no line)
4: library_main +5297955 bytes (no line)
5: library_main +371112 bytes (no line)
6: library_main +3400005 bytes (no line)
7: library_main +3400443 bytes (no line)
8: BaseThreadInitThunk +36 bytes (no line)
9: RtlGetAppContainerNamedObjectPath +311 bytes (no line)
10: RtlGetAppContainerNamedObjectPath +263 bytes (no line)

bertro's picture
User offline. Last seen 2 days 13 hours ago. Offline
Joined: 05/02/2011
Groups: Beta Folders
Basically same UNHANDLED EXCEPTION:

2 times now, doing nothing, client maximized, latest devprev, win10 pro

1: alIsAuxiliaryEffectSlot +9025020 bytes (no line)
2: alIsAuxiliaryEffectSlot +598984 bytes (no line)
3: library_main +5297766 bytes (no line)
4: library_main +5297955 bytes (no line)
5: library_main +371112 bytes (no line)
6: library_main +3400005 bytes (no line)
7: library_main +3400443 bytes (no line)
8: BaseThreadInitThunk +36 bytes (no line)
9: RtlGetAppContainerNamedObjectPath +311 bytes (no line)
10: RtlGetAppContainerNamedObjectPath +263 bytes (no line)

alcor29's picture
User offline. Last seen 14 hours 54 min ago. Offline
Joined: 11/16/2012
see feedback
Joined: 06/24/2008
Groups: Void Crushers
crashed

Created should have been crashed

Joined: 06/24/2008
Groups: Void Crushers
Scoring Strangeness

My score on the board is 13562; my score on my saved/shared solution is 13577 in big type at the top.
In small type underneath my name the score is 13557.

I have also created about three times doing various things.

1519 scoring weirdness1519 scoring weirdness

jeff101's picture
User offline. Last seen 16 hours 15 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
Still active but worth 0 points now:

It looks like because of all the crashes on this puzzle,
you decided to make it worth 0 points but keep it active.
Will we be able to load our solutions from 1519 into the
next version of 1519?

toshiue's picture
User offline. Last seen 9 hours 40 min ago. Offline
Joined: 01/31/2016
Groups: Go Science
0 points, and seconding jeff101's question about 1519 sequel

Hoping that we get to carry over what was a great deal of work into the successor to 1519. Would also like to continue work on the zero point game, and would like to ask if the lab would consider extending the expiration date so we could work on it more slowly around the other puzzles. Need more time to learn that ligand tool. Thanks.

jeff101's picture
User offline. Last seen 16 hours 15 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
LUA function structure.GetAtom to return atom types

At least one person (besides me) has mentioned
the Feedback below regarding this puzzle:
https://fold.it/portal/node/997504

Joined: 12/27/2012
Groups: Beta Folders
Video

Video showing how to build TPV...

https://www.youtube.com/watch?v=qQIdtGQLDmI

Joined: 05/19/2009
Groups: Contenders
MMFF Wiggle ?

What does MMFF Wiggle stand for ? I noticed it on the classic interface 'cartwheel' selection.

S0ckrates's picture
User offline. Last seen 22 hours 57 min ago. Offline
Joined: 05/19/2017
Groups: None
2 Answers

From the forum thread on toxicophore detection actually being ideality issues, rmoretti writes:
"The MMFF wiggle (either in selection mode or from the right-click pie menu) should also greatly help with ideality issues, though as it's working with a different energy function than the normal wiggle, there's a chance an MMFF wiggle will decrease your score slightly."

From Loci's Tips and Tricks Thread:
"Sometimes the molecule looks squashed, with distorted rings and other problems. The MMFF (Merck Molecular Force Field) wiggle tool can help. The ligand must be selected to use MMFF wiggle"

rmoretti's picture
User offline. Last seen 1 hour 30 min ago. Offline
Joined: 01/15/2010
Groups: None
Merk Molecular Force Field

"MMFF" stands for "Merk Molecular Force Field". It's a different scoring function tailored for small molecules.

Sometimes the small molecule design tools can make a small molecule that has bad internal geometry. (This will show up as really bad "Ideality" scores in the TAB residue information menu.) The MMFF wiggle can be used to attempt to remove these bad ligand geometries. But as it's a different scoring scheme than the normal Foldit one, it may actually cause your score to drop.

Regular wiggles can also help with some of these geometry issues. Global wiggles help a little, but don't do too much with really poor geometries. Local wiggles of just the designable small molecule (like the wiggle in the 'cartwheel' menu in the classic interface) use a slightly different approach which works much better in this regard, about as well as the MMFF wiggle.

You might want to try both MMFF wiggle and the single-residue wiggle, and see which works best for your structure. Given that they use slightly different (but related) scoring schemes, alternating between them may (or may not) help wiggle the structure out of "stuck" conformation.

jeff101's picture
User offline. Last seen 16 hours 15 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
MMFF questions:

I found the MMFF wiggle button in the selection interface.
It appeared as a button labeled "M" when I selected just
the ligand. When I selected the ligand plus parts of the
protein, this button disappeared. Is it possible to do
MMFF wiggle on more than just the ligand?

In the original interface, why not put the MMFF wiggle
button in the Actions menu along with all the other
wiggle buttons?

Are there hot-keys or LUA functions for MMFF wiggle?

Joined: 05/19/2009
Groups: Contenders
Awesome, my score just

Awesome, my score just jumped. Thank you !

frood66's picture
User offline. Last seen 48 min 51 sec ago. Offline
Joined: 09/20/2011
Groups: Marvin's bunch
This is designing drugs -

This is designing drugs - nothing to do with folding....this should not be in this game.

create another one....I'm sure that those interested will try it.

Foldit is about folding proteins - please remove this puzzle from this place

:)

S0ckrates's picture
User offline. Last seen 22 hours 57 min ago. Offline
Joined: 05/19/2017
Groups: None
They go hand in hand

To me this puzzle is Aflatoxin reversed; instead of designing the protein, why not design the small molecule? It's a clever way to find new ways to play the game. Also for the record, protein based drugs (or biologics, I think the term is) are a thing too, just not what you'd expect when you say the word "drugs." There's still sidechain manipulation involved, so it's not like folding is out of the question. I'd argue that understanding folding is needed to sort of "level up" to these kinds of puzzles.

To me, there is nothing strictly dictating that a game shouldn't evolve and expand over time to reach new horizons and reach across disciplines. I come back to games that get updated with new features to check them out, and sometimes I end up loving the game all over again. Of course, full disclosure: I'm totally biased since I myself am getting into pharmacy school. But to me, this serves as a really cool opportunity to harness something existing and make something really cool from it. Modding games have turned other games into completely new genres different from the base game, and these mods are loved just as much as base games. Who's to say the developers can't mod their own game with new expansions and game modes?

Joined: 12/27/2012
Groups: Beta Folders
molecular weight penalty

A little late, but for molecular weights over 630, the penalty is 15 points per g/mol.

So a weight of 630.702 gets a penalty of 10.53 points, while a weight of 704.869 gets a penalty 1123.04 points.

Just for future reference....

Get Started: Download
  Windows    OSX    Linux  
Windows
(7/8/10)
OSX
(10.7 or later)
Linux
(64-bit)

Are you new to Foldit? Click here.

Are you a student? Click here.

Are you an educator? Click here.
Search
Only search fold.it
Recommend Foldit
User login
Soloists
Evolvers
Groups
Topics
Top New Users
Sitemap

Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons