I just realized it was 10 years ago that I started this blog. I should start writing again
The More the Merrier: Combine Drugs Together
256th ACS National Meeting
Boston, August 19-23, 2018
Dear Colleagues, we are organizing a symposium at the Fall ACS meeting in Boston focusing on computational, experimental and hybrid approaches to investigate the effect of drugs combinations on biological systems and characterizing phenomena such as synergy and antagonism. The aim of this symposium is to explore fundamental principles that underlie effective combination treatments and synergistic drug behavior across different diseases; to reveal the possible side effects of combination treatment; to investigate possible mechanisms of drugs synergy, additivity and antagonism; to develop a new computational approaches for prediction of these effects.
We welcome contributions related to the identification of promising or novel drug combinations using to treat the rare/neglected disease as well as socially significant diseases (cancer, HIV, etc.), approaches to characterizing drug synergy scores and novel statistical and machine learning methods to predict and prioritize drug combinations as well as to identify side effects caused by drug mixtures.
The deadline for abstract submissions is March 20, 2018. All abstracts should be submitted via MAPS. If you have any questions feel free to contact Alexey or myself.
Move Away from the Lamp Post & Find Druggable Targets
Dear colleagues, we are organizing a symposium focusing on methodologies and case studies that have explored under- or unstudied targets, with the goal of elucidating their function or role in the context of human disease. Several reports have highlighted the focus on current biomedical research on a relatively small set of protein targets. For example, Edwards et al (Nature, 2011) reported that 75% of research is focused on studying only 10% of the known mammalian proteins.
Clearly, there may be many unexplored opportunities among the set of under- and unstudied targets, the so called “dark targets”. Our own report (Nature Rev Drug Discov 2018) suggests that as many as 38% of human proteins are significantly understudied, and that over 9000 human proteins are currently not associated with NIH funding.
At this critical juncture in human health research, we believe it is timely to discuss approaches to study the “dark genome”. We invite submissions that address all aspects of illuminating dark targets, with a particular focus on computational approaches that involve novel experimental methods. Topics of interest include, but are not limited to:
- Target prioritization methods
- Integrative approaches that combine data types, using well studied targets to shed light on unstudied targets
- Case studies that have elucidated function or disease relevance for unstudied targets
- Novel characterization of targets that go beyond traditional structure based approaches
- Target-centric databases that highlight dark targets
The deadline for abstract submissions is March 20, 2018. All abstracts should be submitted via MAPS at https://acsnm256.abstractcentral.com/. If you have any questions feel free to contact Tudor or myself
The Chemical Structure Association (CSA) Trust Grant Program is now accepting applications, from young researchers with an interest in systems and methods used to store, process and retrieve information about chemical structures, reactions and compounds. You can view the details here and can get in touch with Bonnie Lawlor (firstname.lastname@example.org) if you have any questions.
Recently Barbara Zdrazil and I published an article that explored the idea of tracking the attention being paid to a scaffold in the medicinal chemistry literature (as represented by ChEMBL). The gist of the idea is that scaffolds that are more frequently enumerated or tested in more assays (or even published in increasingly high IF journals) are receiving more attention than ones that are less frequently enumerated and so on. By fitting robust regression models to per-year scaffold-aggregated properties we identified significant vs non-significant trends.
The idea originated from a blog post (archived version) by Jonathan Baell, where he traced the publication history of the bis-chalcone scaffold starting from Stoll et al, Biochemistry, 2001 ending up at Anchoori et al, Cancer Cell, 2013, the point being that a PAINS containing scaffold (and thus of possibly dubious biological activity) received increasing attention resulting in a (relatively) high profile journal publication. This led to the question of whether we could systematically capture such attention trends for other scaffolds and thus this paper.
While the article presents a comprehensive analysis, it is limited to using a fixed set of scaffolds (defined using the Bemis-Murcko scheme) and a few properties we selected because we thought they would be proxies of attention. What if we could consider any scaffold? And visualize the time evolution of an arbitrary scaffold-aggregated property over time? This would be something like Google Trends – except that instead of text search terms, you’d be comparing scaffolds.
So I put together the Scaffold Trend Explorer, which allows you to view trends for a number of properties, for arbitrary substructures. Obviously, searching for frequent substructures will probably be non-responsive (so I disallow queries such as benzene and straight chain alkanes with < 8 carbons). I’ve provided a number of properties ranging from the count of enumerated compounds to drug-likeness. You can draw a structure or provide the SMILES directly. In addition there is a set of bookmarks for well known scaffolds (taken from Welsch et al, 2010). You can compare multiple (up to 9) scaffolds at a time, and compute moving window average curves, which hides the year to year variation.
This tool should let users play around with the idea of scaffold trends. Currently, it’s a very simple visualization tool – you can download the per-year data, but that’s it. Unlike the paper, I don’t fit regression lines, though I hope to implement this in the future. There’s a number of enhancements planned, including access to the underlying publications for a scaffold in a given year, simple analytics (such as differential analysis) on trends and so on. A natural next step is to go beyond the medchem literature and consider patents as well (say, via SureChEMBL). And of course, feature requests are also welcome.