October 29, 2018 Joint Governing/Advisory Board Meeting
- Currently we have eight partners on board (with openforcefield.org listing those whose legal departments have OKd disclosure)
- Two additional companies
- Two additional companies are awaiting funds
- Two are working on finalizing approvals
- David Mobley is following up with several others
- Invoices to go out this week from MolSSI to those already on board
- The board discussed approving use of funds for discretionary funds and routine expenses below a particular threshold without a full vote
- Travel support for investigators was approved as needed
- The board decided to begin advertising someone to work on a tool for bespoke/in-house parameterization of small molecules within the framework of the initiative
- Other potential hires: torsion refitting, electrostatics fitting with Psi4, automated benchmarking were discussed. To be revisited as additional funding comes online.
- We discussed the scheduled face to face meeting in San Diego tentatively 1⁄7 and 1/8/2019:
- PIs to put together draft agenda
- Possibility of 2-day meeting discussed, seemed to be the preference
- Potential for 2 morning half day sessions broadcast by Zoom for highest priority items of general interest to the advisory board followed by more in-depth breakout groups in afternoons
- Allows participation from Europe
- Final decision will be made about 1 full day vs. 2-day meeting after agenda drafted
- Possibility of 2nd face to face meeting around San Diego ACS to be investigated – seemed to be a decent possibility for many. We will plan on the Friday and Saturday after ACS, which wraps on Thursday.
- Full Zoom meeting in between to be discussed at a later date
- Advisory board meetings will be monthly
- Use Github issue tracker to report problems, advisory board members to sign up/send usernames to receive prioritization
- Science updates to be posted on website (subject to prioritization)
- Governance board meeting summarized minutes to be posted on Consortium page
- Software: Wagner to set up 1:1 meetings to understand specific needs for partners
Attending: All PIs, both governing board members, and most advisory board members, via Zoom.
Short-term roadmap/minimum viable product:
In general folks are supportive of current effort plans. However, there was consensus we need a tool for bespoke parameterization of molecules, especially internal molecules which might not be able to be disclosed to the larger effort, as part of our minimum viable product.
We discussed how to work with industry to prioritize particular chemistry which might be proprietary. There is the possibility for partners to identify public molecules with fragments of interest for parameterization. Clara Christ and others suggested using molecules from disclosed patents to enrich pharmaceutically-relevant chemical space using, e.g. SureChEMBL.
There is broad support from advisory board for RDKit integration as part of the ultimate OpenFF goals. There were questions, however on cost/benefit of near-term RDKit integration in the full set of software: - A clear priority is RDKit support in the public-facing tools/tools for using force field. RDkit support elsewhere is subject to prioritization - All tools need to be fully open eventually