Personally I'd love to build such a system, I'm just swamped right now with a million and one projects. Any other engineers in here?
A wiki like this is a fantastic idea. I'm definitely interested and I've got some time now to work on it.
Great! If you'd like to help out, then just pick whatever compound your interested in and share a link to it through workflowy or dropbox.
I consider the stuff I post on workflowy to be much more refined, but I've got a bunch of less organized notes in this dropbox folder.
https://www.dropbox....cdqd/xVVfA7jMgq
If this thing starts taking off, perhaps we could start using a facebook group, a mailing list, or just listing everybody's contact info in workflowy to keep everyone motivated to keep working.
How is it that this post on Cerebrolysin has 200,000+ views whereas Chrono's Nootropics Thread Index only has 64,000+ views?
^ Google. All about SEO, and I bring that up because our SERP rankings on various names (layman, research, generic, brand) will affect the user-base and by corollary the quality of source, discussion, and recency.
I'm a designer/artist turned engineer--I believe function, in a sense, is molded by form and that form is the immediate impression which frames the experience. All that not to be lofty, but to say that ultimately two factors will determine success:
1) Usability; in the sense of researchers' ability to easily contribute, and more importantly, receive some kind of value (either extrinsic or intrinsic) for improving the community. What this 'value' is should be determined by someone actively involved in the field of chemical/pharmacokinetic research as they will likely have insight into the professional motives of similar individuals, and more importantly, will know how to keep out those undesirable (big pharma, biased and/or sub-par research). Making sure the right people have incentive to contribute quality information makes the community self-sustaining, and prevents the inevitable intellectual decline you see with mediums like web forums (especially when they undergo rapid growth; reddit is a good example though a different context entirely).
2) Clarity; Information needs to be laid out PRACTICALLY. I've known many people from Ivy League research communities and a consistent trend seems to be the desire to "archive" information. So much time is spent trying to figure out how to classify and store information that the bigger picture and more abstract connections are often lost. Now we all know that proper classification of data is a critical precept to quality research, my point is that the only way for this to gain any traction in a broader demographic is to feature the *practical implications* of this data in a context that is accessible to those without a degree in biochem (all the while not alienating those interested in more formal analysis)
I can provide this second perspective, as I have an avid interest in pharmacokinetics but have never received any formal instruction in the field. As for the first, we need someone experienced in the research community to weigh in. I can draft the ideal front-end UI, but the more complex features catering to these content providers will have to be determined by someone with appropriate perspective in the field... We need a blend of these archetypes to create an optimal UX.
I think the ideal outcome is something that comprises the best qualities of the following platforms:
- Wikipedia: seamless collaborative editing, and bibliographic capability (note: their reference system is actually horribly designed--needs a serious usability overhaul)
- Chemspider: easily accessible documenting of physical properties, patents, chemical data (also a very basic UI but data is accessible which is a UX plus)
- Erowid: low quality but high volume of anecdotal evidence and a 'guinea pig' magnet (with formatting guidelines, quality could improve slightly and with enough volume researchers could search for commonalities [side effects, etc] to develop hypotheses on bleeding edge research before receiving clinical trial approval)
Wish I had the time to spearhead development, but it's just not feasible for me right now. If someone comes along who's seriously motivated to do this right, I'm happy to guide them on the implementation.
That said, personally I'd rather see it done right or not at all; I know this might come off as elitist, but as someone who's made it in the tech world, I've seen a lot of sub-par developers waste everyones time trying to justify things that don't qualify as *practical solutions*. Another chem database just won't cut it, unless you're looking for a myopic sampling of information and ~1k/mo from AdSense.
Bottom line, this is an original idea that doesn't currently exist in the capacity we've envisioned. If it's built *correctly*, it will become an authority database to professional researchers, students, and even the general public in exponential fashion.
[edit] Another brainstorming point but I seriously have to get back to work... The discussion/wiki area should probably have a simple & subtle color scheme to easily identify the trust level of various pieces of information. A simple algorithm could calculate this based on the # of sources & their respective trust ratings as determined by a moderator/researcher/medical authority/etc. This way everyone can efficiently locate information relevant to them without us having to segregate it. Of course, there should still be in-page relative links to more detailed breakdowns that are comprehensive and tailored to each type of user.
Also on that note, modularity is important--everything needs function seamlessly to provide experiential simplicity. If you're looking at a research study, it should show relevant & ranked information for all related fields (compounds, researchers, follow-up/chronological studies, assertions, and relational links [bio. pathways, qualitative info, pkd background for laymen]). Similarly if you're looking at a compound, all related components need to be featured (studies, documentation, mechanism(s) of action, experiences, discussion). These all need to be relational, and seamlessly in sync not just on a static database level, but qualitatively--it's not just display of information, but trying to make that information as *useful* as possible and potentially provide insight.
Edited by lmlj, 15 October 2012 - 07:30 AM.
