mattodd
commented
6 years ago Hi @mcoster. A great write-up of your recent experiences.
We have what I think is an important resource (that you've added to) in the DCNYS page. I had experimented with the same idea for methodology, essentially mini-projects in relevant, required methodology that might be attractive to people coming to OSM. We should probably install such things on the DCNYS page, perhaps in a separate section on synthetic branches, just to keep everything together in a chocolate box of targets. This is relevant to your point about a residual project for new undergrads next time around. An alternative is that we use the organic chemistry section of the relevant wiki for this, assembling current and past attempts to solve methods and highlighting gaps, e.g. like this.
However, more generally your plan for an UG project investigating the core synthesis would be absolutely fantastic. The need to repeat the early steps of the synthesis is a little slowing, yes, although in my experience it can help students settle in a new lab. @cdsouthan was asking around for companies who might be interested in some bulk synthesis of intermediates (OpenSourceMalaria/OSM_To_Do_List#369), but no dice yet, I think. The outputs of your project could fill that gap if the projects worked well.
I agree with you that direct engagement of junior students with Github can be an issue. It's not clear to me why.
Freely-available chemical drawing packages can be sub-optimal. I should re-engage ChemAxon (OpenSourceMalaria/OSM_To_Do_List#531). But any in-roads into something lightweight for drawing structures for ELNs or Github pages would be very nice. You say "no cheminformatics" but if we were able to use the software to generate (png, jpg) images that contained metadata that included SMILES that would go a long way to machine readability with minimal human hassle. I'm sure lots of people would be interested in a beta test whatever.
For the Trimester 1, 2018 projects then, yes, the DCNYS targets are all good to go, though I guess there is now some interest in David Smith's collected analogs of the Pfizer compound (OpenSourceMalaria/OSM_To_Do_List#555)
Re your comments in C) about cognitive load - you're addressing significant barriers here, so many people will be interested if you find solutions. The nature of the tools we use is secondary to the idea that all data are recorded, are public domain, have provenance and permanence and are ideally machine readable. For videos (great idea) we could host those on the Tech Ops repo/wiki.
Newsletters. I'm about to send out the first, just because it's all done. Just commented on this: (OpenSourceMalaria/OSM_To_Do_List#547).
Series 4 progress
During trimester 2, 2017, six Griffith University undergraduate students took on OSM projects for credit towards their degrees. Planning and discussion can be found in these OSM_To_Do_List Repository Issues:
What did we learn?
Unfortunately, we did not produce any new molecules for biological testing. As a new research area, it also took us some time to get "used to" the synthetic chemistry. We actually spent considerable time and effort investigating the first few reactions with a view to streamlining what are already quite straightforward procedures. This has inspired us to develop the synthesis of a key triazolopyrazine intermediate as a core undergraduate experiment (details below). We will report the "undergraduate lab experience" optimised procedure in due course. Apart from that, we learned the following:
Future Work
We're very interested in continuing as OSM contributors, in an expanded fashion. I think we've got past some of the teething problems and we're in a better position to plan our future contributions.
A. Triazolopyrazine - undergraduate synthesis experiment
I convene a third year organic chemistry course that includes 6 x 4 hour lab sessions, covering two organic synthesis experiments. One of these (introduced many years ago) aims to expose the students to a number of chromatographic techniques, both preparative and analytical, including a comparison of recrystallisation and flash chromatography as purification techniques(purified material analysed by GC). I have been dissatisfied with this experiment for a number of reasons: 1) the experiment is "recipe-following", ie. students do not monitor the reaction, but reflux for a pre-determined period of time 2) the flash column is conducted in an inauthentic way - it uses alumina and there is no attempt to fractionate 3) the product material is of no use to us and is sent for disposal at the end of the trimester
It struck me that the triazolopyrazine synthesis would be an excellent replacement for this venerable experiment. The chemistry is robust, reactions are readily followed by TLC, products can be purified by recrystallisation or flash chromatography on silica gel and the product materials are valuable intermediates in an Open Source drug discovery initiative! Hence, we are developing this as a new 3rd year undergrad experiment for introduction Trimester 1, 2018. Our main areas for development are making sure the reactions go to completion within a suitable timeframe and that the procedures are robust using the equipment and resources available in the undergraduate lab. We will report on progress. We hope to use the product from the undergraduates in future OSM projects and to make available any excess to other OSM contributors.
B. Trimester 1, 2018 Projects
There will be ~ 8 new undergraduate students at Griffith Uni, taking on OSM projects. They will be defining their individual goals ~ mid-late February. At this stage, I suspect we will continue to target some of the DCNYS compounds, plus other compound(s) of more recent interest. We're open to any suggestions and will engage in consultation via GH issues shortly.
C. Minimising cognitive load on project students
The cognitive load on students taking OSM projects with me is very high. Most of these students are studying at Nanjing University of Traditional Chinese Medicine and only articulated into our third year program 6 months prior. They are learning new synthetic chemistry techniques, research skills, scientific writing skills and project management. They are new to Scifinder, have not kept a research labbook (electronic or paper) before and have never heard of GitHub. The potential for overwhelm is very significant. To minimise this, we are looking at:
D. cDraw
My colleague, A/Prof. Andreas Hofmann, and I have been developing (ok,.. Andy is doing all the coding!) a very lightweight chemical structure drawing program, currently called cDraw (but, I also like YABCSDP - Yet Another Bloody Chemical Structure Drawing Program). It is in the early testing at the moment with hopes of beta testing early next year for anyone interested, plus my undergraduate students. The focus is on:
It is Java-based but with no dependencies, weighs in at ~ 560 kB download. The hope is that it may prove useful to chemists who just need to draw presentation quality structures and schemes with minimal fuss and at no expense. I will put out a GitHub Issue when it's ready for external testing.
E. OSM Newsletter
As previously discussed, we have mocked up a Flat-file CMS blog-style Newsletter and recent developments site - http://test.coster.group/ Happy to roll out as a production site if this is an attractive option for OSM.
(Note: I may edit this further - treat as draft. Feedback welcome.)