Foldit drug design introduction

Hello everyone,

My name is Steven Combs (aka free_radical). I am currently a post doc with a dual appointment at Vanderbilt University and Eli Lilly. I have been working with David Baker’s lab and the developers of Foldit to enable drug design in Foldit.

During one of the developers chats, it was mentioned that players wanted more updates on new developments in Foldit. I will try and update everyone as much as possible on my progress for drug design in Foldit and explain some of the scientific ideas behind the implementations in the game.

To start off, I would like to explain one component that has changed in Rosetta (the underlying software for Foldit) to enable drug design. Rosetta assigns properties to atoms based on the type of atom. These properties can be anything from whether the atom is a hydrogen bond donor/acceptor to whether the atom likes to be exposed to water or not. Further, numerical values used in scoring a residue based on its atoms can be assigned. Many of these values used in scoring are derived from the CHARMM force field, which was developed by Dr. Karplus (who just recently received a Nobel Prize in chemistry!).

While these values help with scoring the residue and atoms, they do little to tell about the configuration of the atom in relationship to other atoms bonded to it. This is extremely important in drug design. For drug design, the type of bonds that can be added or deleted or the types atoms that can be added or deleted need to know what the configuration of the original atom was. For example, if an atom is double bonded to another atom, can that atom form a triple bond? Does it have any free electrons to participate in another interaction? When building small-molecules for drug design, these properties, or chemical rules, need to be known.

To do this, I, along with members from the Meiler lab, have worked to put new atom types into Rosetta. I will use the amino acid TYR as an example of the new atom types. Below is a diagram of TYR with some of the atoms labeled with their properties assigned by Rosetta using the old atom type scheme.


Several properties are encoded onto the atom, such as the carbon being aromatic and the oxygen being polar. These properties are very useful when scoring the side-chain, but we also need to add on a layer for encoding the configuration of the atom.

The rules that we use to encode the configuration are based on the geometrical configuration of the atoms in relationship to what is bonded to and the number of electrons in the bonds (referred to as Gasteiger atom types). For our TYR example, the aroC retains the same original properties, but we also now know its geometry.


The new atom type is C_TrTrTrPi. This means that the carbon has three bonds that are in the trigonal configuration. Trigonal configuration refers the VSEPR rules. The Pi at the end of the naming means that there is one pi-orbital in the system, occupied by one electron. That pi-orbital is free to interact with other hydrogens or other pi-orbitals to form a cation-pi interaction or pi-pi interaction, all which are important for drug design (more on this topic in the future). For the oxygen, it is now labeled O_Te2Te2TeTe. This means that there are two lone pairs in tetrahedral (sp3, Te2Te2) and two bonds in tetrahedral configuration (TeTe).

While amino acids will not see much use for these types of descriptors for drug design, small molecules will. For example, lets look at a cyano group, which is a common group used in drug design.


In the cyano group, the old Rosetta designation for the atom is aroC, but the configuration of that atom is much different than the aroC seen in TYR! If we were to modify the atom, how would we know the configuration of the bonds? This is where the power of the new atom type comes into play. With the new atom typing, we now know that the carbon is linear (the DiDi portion; Di=diganol/linear) and that it has two pi-orbitals (PiPi). This means if we add or replace atoms, we know exactly the placement for the new atoms and the type of interactions this atom can make.


While these modifications may seem small, they greatly enhance the ability of Rosetta for drug design. With the new atom types, we can combine/add/delete/modify residues and small molecules rapidly and with ease.

For the upcoming weeks, are there specific topics that you would like to be addressed? What would everyone like to hear about? If anyone has any questions on this subject, I will be more than glad to address them!

( Posted by  free_radical 164 15516  |  Wed, 01/21/2015 - 17:41  |  11 comments )
Susume's picture
User offline. Last seen 4 hours 40 min ago. Offline
Joined: 10/02/2011

Thanks for the update!

I am confused about the pictures of the PHE. With the OH tip, it looks more like a TYR to me.

Still, exciting to know we can have small molecules soon!

Joined: 01/12/2015
Groups: None
you are correct. The picture

you are correct. The picture is of a TYR. I will go back and fix that!

Joined: 09/24/2012
Groups: Go Science
What I understand

That originally, Roseta used a simple info about the sidechains bonds: only polar, donor, acceptor, aromatic ...
Now you know more about the tridimensional configuration for each aminoacid and you want to implement it into the info available for scoring.
This allows to predict some potential for bonding or not.
Then, in the futire, the scoring function of Foldit will be better for better medecine designs.
Is it correct?

(btw: I suppose you mean trigonal configuration? not triganol?)

I dont' see the link between your picture and the wiki trigonal planar pictures.

I don't understand the pi-orbital stuff.

Thanks for trying to give such info! may be a link to some other simple wiki or course?

Joined: 08/30/2011
'I dont' see the link between

'I dont' see the link between your picture and the wiki trigonal planar pictures.'

Tirgonal planar means that there is a bond angle of roughly 120°. In our case we were looking at a C-atom in a aromatic ring. This specific C-atom has 3 binding partners: 2 other C-atoms and 1 H-atom. The most favorable configuration for an atom with three binding partners is the trigonal planar configuration.
The electrons sit in between the atoms and they repulse.
With 4 binding partners the electron repulsion leads to a tetrahedral geometry.

'I don't understand the pi-orbital stuff.'

I really try to keep it simple and short.
There are several types of orbitals in an atom. If they are close together, these orbitals can be combined. And if you combine 2 orbitals you will get a set of 2 new orbitals. One of these new orbitals is the so called bonding orbital. Its energy is lower. The other orbitals energy is much higher than before and is therefore called antibonding. If you would put electrons in this orbital, you would get a pretty unstable molecule.
The gap of the bonding and antibonding orbital is determined by how good the two initial orbitals overlap.

The more orbitals you have, the more complicated is their structure (resulting in Pi orbitals). So the overlap of the 'higher order' orbitals isn't that good. Resulting in a overlap which isn't that big. This means if you combine these you will get a new set of orbitals which don't split up that much. You will get some stabilisation which is very favorable. BUT the stabilisation energy isn't that big like with the simple orbitals.
There are systems in which it makes sense to put electrons in Pi-antibonding orbitals, because the overall gain in stabilisation energy is bigger than the destabilisation it generates (because it is antibonding).

I hope I could help a bit. This whole orbital theory can get really really confusing. Some people consider this university level and not highschool level anymore.

LociOiling's picture
User offline. Last seen 1 hour 17 min ago. Offline
Joined: 12/27/2012
Groups: Beta Folders
more info for the perplexed

I was deflected by the "CHARMM force field" and trying to follow the "VESPR rules", but then Susume posted this useful link in veteran:

If nothing else, it may help a bit with the next electron density puzzle (if anything can...).

Joined: 08/30/2011
exciting news

Thanks for sharing!

I am really excited about this and can't wait to design small molecules. I really hope it will be open for public soon.

It's VSEPR and not VESPR.

spmm's picture
User offline. Last seen 8 hours 55 min ago. Offline
Joined: 08/05/2010
Groups: Void Crushers

If I look at a Lewis diagram of amino acids which has dots representing the free electrons which will bond, this one doesn't have dots for some reason :(

So I think you have implemented a way of scoring the '3D orientation' and strength of these bonds (in a small molecule) against a set of rules in a 'folded up' space as well as retaining the other rules we (mostly) know already.
Is that in the ballpark or am I way off?
Looking forward to giving these a try and keen to see the game space. And checking how they wiggle :P

jeff101's picture
User offline. Last seen 3 hours 53 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
Dots in Lewis Structures:

Often single bonds - and : are used interchangeably.
Also, double bonds = and :: are user interchangeably.
Triple bonds would have ::: (6 dots).
These are because each covalent bond holds 2 electrons (2 dots).

Lone pairs contain 2 dots. For example, in NH3:


the N has one lone pair. Also, in H2O:


the O has 2 lone pairs. The lone pairs occupy space, so they distort the geometry, making NH3 trigonal pyramidal and H2O bent. gives some other examples.

The dots matter more when things take on free radical properties (like paramagnetism).

Joined: 05/19/2009
Groups: Contenders
Computing performance

What requirements can we expect for the upcoming software release ? Does it use GPUs, how big are the models expected to be ? How big can the models get ? I ask because of upcoming hardware purchases.

Thank you for posting this, love to read more.
And thank you to the other contributors to the thread, I am impressed.

jeff101's picture
User offline. Last seen 3 hours 53 min ago. Offline
Joined: 04/20/2012
Groups: Go Science
Will Free Radical Reactions be possible in Foldit?

"The Pi at the end of the naming means that there is one pi-orbital in the system, occupied by one electron."

Does this mean the new system will be able to handle free radicals (molecules with unpaired electrons)? For example, nitric oxide (NO) and superoxide (O2-) are two very important biological free radicals, and free radicals can be very reactive. Will free radical properties be able to migrate from one molecule to another? Will Foldit let homolytic cleavage occur (where a single covalent bond breaks to form two free radicals)? Will Foldit let radical-radical recombination occur (where two free radicals combine to form a single covalent bond)?

Will transition metals be treated similarly? They can have unpaired electrons in their d orbitals and so can behave like free radicals.

Joined: 01/12/2015
Groups: None
Answer to questions

Hi Guys,

Sorry for the delay in response. I have been working on trying to get Foldit to build for Windows using a different compiling system. We are doing this in order to get some of the code from the BCL into Foldit for windows users.

@susume, @mottiger and @bruno
I am a notoriously bad speller and constantly make mistakes with acronyms. I remember when I was in first grade, I failed a spelling exam because I couldn't spell the word 'spell'. I have lab members and Nova check my work, but some things still escape :D

@bruno, @spmm
Currently, we are just using the new atom types for describing the configuration of atoms algorithmically. This is really important because, like shown in the figures above, ligands can have a wide variety of configurations that are not described with the Rosetta atom types. In the future, we do plan to add partial orbital charges to the new atoms, but scoring functions are really tricky. As everyone who has lived through the switch in Foldit from the original Rosetta score12 score function to the new talaris2013 score function can attest to this fact. The way we score the configurations will be with the Foldit score function, but the orientation will be described by the atom types. Spmm hits on an interesting point with wiggling the small molecule. Next week there will be a blog post on how we implemented the wiggle moves!

Hybridization of orbitals help describe the 3D space occupied by the atoms and bonds between the atoms. There is a lot of chemistry and mathematics that goes into this definition, but the easiest way to describe it is that electrons are repulsive in nature. They like to be the furthest away from each other. When they interact with other electrons they form hybridizations. Its some really cool stuff!

@Bletchley Park
The algorithms that we are implementing should not require more powerful computers than what you run currently with Foldit. We are very conscious of the fact that people do not have sicentific computers, so we have modified some of our algorithms to run faster. In fact, the original implementation we made for wiggle with small molecules was changed. It was too slow and we wanted to make sure it ran on all computers, so we implemented a real cool algorithm to make it faster (for small molecules, compared to our first attempt). The models will remain around the same size and complexity, but I would be willing to put larger puzzles up if everyone wanted them. I have been in school more than 85% of my life, so I havent had any money to upgrade my computer system :( All I can recommend is to buy refurbished because its cheaper :D

We are playing around with the best way to go about allowing for design. There will be a long blog post about this in the upcoming weeks, but what I can say now is that there will most likely bee a free design and a reaction based design option. Free design will let you design the small molecule however you want while reaction based will require you to follow a set of synthetic reactions in order to make the small molecule. Like I said, this is all developing now, and I should have a cool blog post with graphics on this soon!

Download links:
  Windows    OSX    Linux  
(10.12 or later)

Are you new to Foldit? Click here.

Are you a student? Click here.

Are you an educator? Click here.
Only search
Recommend Foldit
User login
Top New Users

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