openmolecules.org

 
Home » DataWarrior » Functionality » Generating derivatives of parent structure
Generating derivatives of parent structure [message #2027] Sun, 05 November 2023 10:09 Go to next message
lutek14 is currently offline  lutek14
Messages: 3
Registered: October 2019
Junior Member
Hello,

I appreciate the quality of the program, but require some assistance. I have an original structure and wish to generate derivatives by substituting specific locations by certain residues. Is this achievable? Alternatively, could you recommend any freely available software that can accomplish this task efficiently?

Thanks
Re: Generating derivatives of parent structure [message #2028 is a reply to message #2027] Sun, 05 November 2023 19:14 Go to previous messageGo to next message
nbehrnd is currently offline  nbehrnd
Messages: 211
Registered: June 2019
Senior Member
Hello,

this taps into the generation of a combinatorial library you find below Chemistry -> enumerate combinatorial library. Either you find your reaction among the templates DW ships by default, or you create your own reaction pattern. Tab «reactants» allows you to use molecules already defined in a .dwar, or .sdf; or to let DW select ones which are commercially available (criteria are adjustable).

See chapter «chemical structures», section «combinatorial library» in DW's help or the online web page[1] for additional information. It however is time well invested to follow Isabelle Girault's video tutorial about the topic «RSC CICAG Open Source Tools for Chemistry Workshops:- Advanced DataWarrior»[2] recorded in 2021 around 47:18 min:s on youtube. (There is sequal «DataWarrior workshop by Isabelle Giraud»[3] by her to complement the insight about DW, too.)

Regards,

Norwid


[1] https://openmolecules.org/help/chemistry.html#VirtualLibrari es
[2] https://www.youtube.com/watch?v=mQCf9GakQW0
[3] https://www.youtube.com/watch?v=Is2hLqqSFvM

[Updated on: Sun, 05 November 2023 19:15]

Report message to a moderator

Re: Generating derivatives of parent structure [message #2029 is a reply to message #2028] Sun, 05 November 2023 21:15 Go to previous messageGo to next message
lutek14 is currently offline  lutek14
Messages: 3
Registered: October 2019
Junior Member
Hello, thank you for your response.
However, I am not seeking reactions themselves, but just structures. For instance, I would like to create combinations with substituted hydrogen atoms on the aromatic ring, altered heterocycle, and various sized side chain .


https://i.imgur.com/7ZG9ibf.png
Re: Generating derivatives of parent structure [message #2030 is a reply to message #2029] Sun, 05 November 2023 22:36 Go to previous messageGo to next message
nbehrnd is currently offline  nbehrnd
Messages: 211
Registered: June 2019
Senior Member
The combinatorial library is generated within a .dwar file and eventually contains an array of the structures of the product, as well as as of the corresponding starting materials.

Departing from here, you can either

- edit this array to selectively copy & paste the column(s) of interest / remove columns not suitable, or
- export the .dwar file as such as a .sdf file which then contains only the structures of the products.

See the .dwar and .png attached for a small 10 by 10 library to yield benzophenones as illustration. This approach DW provides is intuitive, the implementation to recur on compounds only known «in house» or commercially available delivers the results quickly.

Regards,

Norwid
Re: Generating derivatives of parent structure [message #2048 is a reply to message #2029] Sat, 09 December 2023 00:42 Go to previous message
thomas is currently offline  thomas
Messages: 679
Registered: June 2014
Senior Member
You could also use the "Build Evolutionary Library..." functionality. This creates new molecules from a starting generation by random small modifications. You can also define parts of the molecule that shall not be modifed by just selecting it. I tried that with your molecule and selected all but 4 ring positions. This causes random modifications at all selected atoms, which includes ring cleavages.

Then you are supposed to add fitness criteria, which are used to value the quality of created structures. Those structures, that match your criteria best will be kept and form at the same time the parent generation for the next round of random changes. I just defined a simple criterion: the molecular weight shall be not higher than 240. Otherwise I just use default options for everything else. The settings look like this:

/forum/index.php?t=getfile&id=792&private=0

In less than one second DataWarrior created 267 different molecules in 15 generations that all contain the selected (red) part of your defined starting structure, with lots of variations regarding substitution, ring size, partially open ring structures and molecular weights always below 240. The picture shows a small subset.

/forum/index.php?t=getfile&id=791&private=0

Of course, you can be more specific with fitness criteria, e.g. use chemical similarities, pharmacophore similaries or even a docking score...
  • Attachment: temp2.png
    (Size: 139.27KB, Downloaded 156 times)
  • Attachment: temp1.png
    (Size: 40.66KB, Downloaded 159 times)

[Updated on: Sat, 09 December 2023 00:46]

Report message to a moderator

Previous Topic: Generate Explanation view based on the data from table
Next Topic: is there a way to continue a macro file on a saved file?
Goto Forum:
  


Current Time: Mon Jul 22 01:36:31 CEST 2024

Total time taken to generate the page: 0.03213 seconds