Christopher Rowley
@rowley.bsky.social
Associate professor in the Department of Chemistry at Carleton University. www.rowleygroup.ca
Training it specifically for a system and using MBIS charges can give us better agreement. These models can also give spurious local minima if they are undertrained.
August 11, 2025 at 5:25 PM
Training it specifically for a system and using MBIS charges can give us better agreement. These models can also give spurious local minima if they are undertrained.
Takeaways: An "out of the box" NNP trained on other molecules will get some features right, but quantitative agreement is lacking for intermolecular interactions, particularly at long range.
August 11, 2025 at 5:22 PM
Takeaways: An "out of the box" NNP trained on other molecules will get some features right, but quantitative agreement is lacking for intermolecular interactions, particularly at long range.
Great work on the prediction of druggable cysteines by group alum Ernest Awoonor-Williams
January 12, 2025 at 4:07 AM
Great work on the prediction of druggable cysteines by group alum Ernest Awoonor-Williams
It was a pleasure to be involved! Fascinating work! I highly recommend the first paper from the thesis to anyone interested in chemical theory and GNNs pubs.rsc.org/en/content/a...
Global geometry of chemical graph neural network representations in terms of chemical moieties - Digital Discovery (RSC Publishing) DOI:10.1039/D3DD00200D
pubs.rsc.org
January 8, 2025 at 6:37 AM
It was a pleasure to be involved! Fascinating work! I highly recommend the first paper from the thesis to anyone interested in chemical theory and GNNs pubs.rsc.org/en/content/a...
Welcome! Keep an eye on the invite codes. We have to bring over the rest of chemtwitter one by one.
October 6, 2023 at 8:57 PM
Welcome! Keep an eye on the invite codes. We have to bring over the rest of chemtwitter one by one.