Search
Monday, 21st of November 2022, 1:07:20 UTC
12:36:10
drmeister
scymtym: Are you online? I am returning to force fields now and the Open Force Field people have made some changes to the SMARTS strings they allow in their force field.
12:36:45
drmeister
There are a couple of SMARTS strings that are causing problems - they look like this...
12:37:21
drmeister
"[*:1]-[#6X4:2]-[#7X3$(*@1-[*]=,:[*][*]=,:[*]@1):3]-[*:4]"
12:39:51
scymtym
if i recall correctly, i went by a specification that is hopefully linked somewhere in the repository
12:40:16
scymtym
but i understand those new strings include extensions to the original specification?
12:40:52
drmeister
Here is the error message I get...
12:41:14
drmeister
https://www.irccloud.com/pastebin/sO88gYED/
12:41:36
drmeister
I don't believe that first error message so I dig deeper ...
12:42:22
scymtym
i'm getting the same error locally for that input string
12:42:32
drmeister
`(chem:parse-smarts "*@1-[*]=,:[*][*]=,:[*]@1")`
12:42:44
drmeister
I think the problem is the `*@1`
12:43:15
drmeister
`@` is a chirality test - I've never seen those outside of `[` `]` braces
12:43:29
drmeister
The `1` is a ring test
12:43:49
drmeister
The `@` may also be a ring bond test.
12:44:41
scymtym
do you know what prompted this new syntax?
12:44:56
scymtym
(compared to the expected use in […])
12:45:11
drmeister
This is their second generation force field - it contains these new strings.
12:45:13
drmeister
https://www.daylight.com/dayhtml/doc/theory/index.pdf
12:46:05
drmeister
There's also this... https://www.daylight.com/dayhtml/doc/theory/theory.smarts.html
12:46:17
drmeister
I dug through them yesterday trying to figure this out.
12:46:30
drmeister
I thought I would ask you - I didn't expect you to get back so quickly - thanks.
12:46:51
drmeister
Next I'll contact the Open Force Field people and try and find someone who can explain what this means.
12:47:09
drmeister
I fear that it's generated by the OpenEye software and it's buried in their code somewhere.
12:47:26
scymtym
asking first would probably be good before changing the grammar
12:58:08
drmeister
I’ll try. My question is does this look like a chiral test or bond test from the parser side.
Monday, 21st of November 2022, 13:07:20 UTC