Back to Recipes Homepage
recipe picture
Recipe: Maaa 2.4.6
4.636365
Your rating: None Average: 4.6 (33 votes)

Profile

Name: Maaa 2.4.6
ID: 101208
Created on: Sun, 07/05/2015 - 09:13
Updated on: Wed, 09/23/2020 - 17:56
Description:

Mutate autorized aminoacid. Quick mutate only to prefered aa types



Best For


Comments

Joined: 09/24/2012
Groups: Go Science
Outlines of the recipe: mutate from and to what you want

It's a kind of quick Mutate Combo that mutates segments one by one, carefully selecting from or to what AA or segment is authorized.

It doesn't try all AAs, carefully selecting what is relevant.

The user can choose to try mutate only to AAs that prefer to be in helixes, or sheets or in the observed secondary structure.

The recipe will not try prohibited AAs either defined in user list, either if it's written in the note of each segment.

After each mutation, different levels of further actions are proposed, from nothing else (mutate only) to share to wiggle. A quick version does not try further action if score after mutate was too low.

A "kill alien" option forces the mutation in order to eliminate an undesired AA (this might cost points).

Joined: 09/24/2012
Groups: Go Science
Used for (start & medium game)

I use it before simple mutate on Design puzzles in order to start with "ideal" AAs before any other action.

I developped and shared an old version of this recipe for L'Alliance francophone in French, then I translated it to English and shared it to Beta Folder.

Not sure they used it a lot.

I added Hbond options and I tried to develop the "select where to work on" section with the aim of beeing able to change one AA type to another. I experienced many bugs trying this, that's why I share it so late to GS. If you see bugs (probably in "select" section), please let me know.

Joined: 09/24/2012
Groups: Go Science
Maaa222

Maaa222Maaa222

Joined: 09/24/2012
Groups: Go Science
2.2.7

Following jeff101's suggestions, added a Probabilistic option.

When checked, it gives more probability of mutation to AAs that are more frequent in proteins.

When checked with "SS out of content", it replaces the simple selection of AAs preferred for each SS (source= wiki Roseta) by a mutation probability based on frequence of AA in Helix, Sheet or Loop.

These probabilities are added to other options. For example, if you select to mutate only to Hbonding AAs, it will only try these AAs but with a probability related to the frequency explained above.

In principle, this new option should speed up the process a little bit, AND the result should be more close to a natural protein than the default "one by one" like recipes.

Joined: 09/24/2012
Groups: Go Science
2.2.8, English and Fran├žais

-Translations are in the function multilingual(lang), currently available in English (default) and French (automatic for players of L'Alliance Francophone and some identified players). Please feel free to translate to other languages.

-Default prohibided are: a g p c (prohibided in sheets and helices) f y w (the ones "scoring too low") - you can edit it in dialog.

-Remark: the recipe code contains useful AA information from various sources. This can be used for further applications within or outside this recipe.

Joined: 03/05/2015
Groups: Gargleblasters
Won't run at all on my latest client

Just a FYI Bruno.

Joined: 09/24/2012
Groups: Go Science
2.2.9

oops sorry ! thanks actiasluna. It's fixed now.

Joined: 09/24/2012
Groups: Go Science
2.3

added option preserve hydrophobicity

Joined: 09/24/2012
Groups: Go Science
2.4.1

Now it should disable filter on request

Joined: 09/24/2012
Groups: Go Science
2.4.2

-- ver 2.4.1 corrected some things in filter management
Updated Stanley to ver 1.2
-- ver 2.4.2: I forgot to enable HBonding !!!!!!!!!!! H-Bonding AAs should work now (faster)

malphis's picture
User offline. Last seen 1 hour 19 min ago. Offline
Joined: 05/25/2020
Groups: None
Attempts to mutate non-mutables

Shouldn't the script automatically exclude non-mutable segments? I started the script on puzzle 1865: Coronavirus Anti-inflammatory Design: Round 8, and noticed that it was attempting to mutate segments 1-151 before stopping it (at segment 35).

After restarting it again, I noticed that by default it tries segments 1-151. I realize that I can specify 92-151 in the options but that really shouldn't be necessary. The output already knows that there are 60 mutable segments, why would it try to mutate non-mutables? It makes me wonder how much computing time has been wasted by the unwary.

BTW, I appreciate the work you've put in on the many, many scripts you've made public!

Joined: 09/24/2012
Groups: Go Science
2.4.3

Thanks malphis.

Updated Stanley to ver 1.2
-- ver 2.4.2 I forgot to enable HBonding !!!!!!!!!!! 17/4/2020
-- ver 2.4.3 changed default selection (mutate only !) thanks to malphis
-- when HBonding puzzle, default not structure prefered (otherwise we have no bonding candidate for helices)

Joined: 09/24/2012
Groups: Go Science
2.4.4

-- ver 2.4.4 Not trying lysine (k) and arginine (r) in Hbond puzzles

(they have too many bonding atoms, there is a high risk of BUNS)

Warning: unlike mutate (that mutate any unfrozen sidechain), this kind of recipe optionally:
-either mutate froozen (sidechain and/or backbone) segments
-either doesn't mutate any segment with backbone OR sidechain frozen.

I had no way to differentiate a frozen or unfrozen sidechain on a frozen backbone. But spvincent found a way. I'll implement it in ver 2.4.6.

Joined: 06/06/2013
Groups: Gargleblasters
Bruno, thanks for option to avoid frozen

This is helping a lot on h bond for folders like me who are not great at mutating. Makes a good recipe even better :)

Joined: 09/24/2012
Groups: Go Science
2.4.6

fixed segment setminus
-- ver 2.4.5 better filter settings 18/9/2020 (avoiding to changing filters when not choosing fast filters), better if you want to work with only one filter enabled (like in HBond puzzles)
-- ver 2.4.6 added selection of unfrozen sidechains
(good for freezing everything unless some sidechains that you want to mutate).
(open menu Select, there uncheck "Don't work on freezed backbones" and select "Don't work on freezed sidechains".

I think that the former version did'nt work on unfreezed backbones only, which could be confusing.

Want to try?
Add to Cookbook!
To download recipes to your cookbook, you need to have the game client running.
Parent
Children
Authors
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