Closed nateharms closed 3 years ago
Merging #85 into master will increase coverage by
0.06%
. The diff coverage is78.94%
.
@@ Coverage Diff @@
## master #85 +/- ##
==========================================
+ Coverage 65.18% 65.24% +0.06%
==========================================
Files 27 27
Lines 4774 4791 +17
==========================================
+ Hits 3112 3126 +14
- Misses 1662 1665 +3
Impacted Files | Coverage Δ | |
---|---|---|
autotst/conformer/systematic.py | 75.11% <78.94%> (+0.61%) |
:arrow_up: |
Continue to review full report at Codecov.
Legend - Click here to learn more
Δ = absolute <relative> (impact)
,ø = not affected
,? = missing data
Powered by Codecov. Last update b90d4d9...10c602b. Read the comment docs.
Hi, could someone take a look at this PR? @rwest @davidfarinajr @skrsna
Sweet, I'm going to merge this in since @skrsna approved.
When I was digging into why AutoTST had a weirdly high failure rate for migration reactions, I dove into some of the results for the conformer analysis and saw conformers that look like the following:
For reaction![rxn](https://user-images.githubusercontent.com/20301547/84808708-15fdb780-afd7-11ea-94c0-6c88e5b7adba.png)
The conformer generation outputted this as one of the low energy conformers where the stars are the reacting atoms:
Diving into this I saw that one of the stared atoms is identified as a chiral center and is therefore inverted during the conformer generation process and this flips the geometry out of its ring-like TS. This PR will ignore chiral centers and CisTrans bonds in rings by looking at the
_pseudo_geometry
which has a fake single bond drawn between reacting atoms. An example of what the_pseudo_geometry
looks like is below.