They are hoping its 2 years away from clinical trials.
I wonder if that's mostly about red tape, regulations and funding, or are there actually two years of actual work to do to synthesize the treatment based on this finding?
Like if a scientist working on the project had terminal cancer and only had a few months to live, would she be able to test this on herself right now?
For what it's worth, I'm not sure about the actual paper itself, but the posted article makes it seem like no one knew that TH588 was not primarily acting as an MTH1-inhibitor.
> “This is an exciting paper for two reasons,” says David Pellman, associate director for basic science at Dana-Farber/Harvard Cancer Center, who was not involved in the study. “First, Yaffe and colleagues make an important advance for the rational design of drug therapy combinations. Second, if you like scientific mysteries, this is a riveting example of molecular sleuthing. A drug that was thought to act in one way is unmasked to work through an entirely different mechanism.”
Since my startup compiles all genomic knowledge from the published literature, of course the first thing I did was search the literature. The first article that comes up with a quick search for MTH1 and TH588 was from 2014 [0], which basically says TH588 is a first-class MTH1-inhibitor effective in treating certain cancer cells. Then in 2016, another article says that's not its primary mechanism of effectiveness [1]:
> In particular, we identified tubulin as the primary target of TH287 and TH588 responsible for the antitumor effects despite the nanomolar MTH1-inhibitory activity in vitro.
Tubulin of course is part of the mitotic spindle which the posted article claims was discovered as the actual target upon which TH588 acts.
I've found around 10 other papers between 2016 and 2018 that further conclude that TH588 doesn't act primarily as an MTH1-inhibitor.
In the abstract sense, I would say we are, in that it's the primary benefit of what we've built versus textual searching like you get with PubMed or Google Scholar. For example, if you were to search for a specific variant like BRAF V600E in Google Scholar or PubMed, you'd need to search for every way that the variant can be described in text, such as p.V600E, p.Val600Glu, c.1798T>A, C1798A, etc. Furthermore, you'd need to then read the results to determine which are specifically referencing the V600E variant in the BRAF gene, versus mentioning the V600E variant in one of the other genes mentioned in the paper.
We've already done that work ahead of time, figuring out which variants belong to which genes in papers, normalizing all the different nomenclatures used for variants so that you can do a single search for your variant and remove the need to think about how authors could have referenced it.
We do the same with taking into account all the different ways authors can talk about genes, diseases, clinical contexts, meanings and interpretations, etc.
I don't know if that answers your question. I guess it depends on what you mean by knowledge graph, since there seem to be a few different ideas of what that entails.
11 comments
[ 3.1 ms ] story [ 34.5 ms ] thread[0]: https://www.cell.com/cell-systems/fulltext/S2405-4712(19)301...
I wonder if that's mostly about red tape, regulations and funding, or are there actually two years of actual work to do to synthesize the treatment based on this finding?
Like if a scientist working on the project had terminal cancer and only had a few months to live, would she be able to test this on herself right now?
A real study would hopefully still be continuing on the normal timeline and the need for it would not be replaced by the act of a desperate person.
> “This is an exciting paper for two reasons,” says David Pellman, associate director for basic science at Dana-Farber/Harvard Cancer Center, who was not involved in the study. “First, Yaffe and colleagues make an important advance for the rational design of drug therapy combinations. Second, if you like scientific mysteries, this is a riveting example of molecular sleuthing. A drug that was thought to act in one way is unmasked to work through an entirely different mechanism.”
Since my startup compiles all genomic knowledge from the published literature, of course the first thing I did was search the literature. The first article that comes up with a quick search for MTH1 and TH588 was from 2014 [0], which basically says TH588 is a first-class MTH1-inhibitor effective in treating certain cancer cells. Then in 2016, another article says that's not its primary mechanism of effectiveness [1]:
> In particular, we identified tubulin as the primary target of TH287 and TH588 responsible for the antitumor effects despite the nanomolar MTH1-inhibitory activity in vitro.
Tubulin of course is part of the mitotic spindle which the posted article claims was discovered as the actual target upon which TH588 acts.
I've found around 10 other papers between 2016 and 2018 that further conclude that TH588 doesn't act primarily as an MTH1-inhibitor.
[0] https://www.ncbi.nlm.nih.gov/pubmed/24695224 [1] https://www.ncbi.nlm.nih.gov/pubmed/27210421
In the abstract sense, I would say we are, in that it's the primary benefit of what we've built versus textual searching like you get with PubMed or Google Scholar. For example, if you were to search for a specific variant like BRAF V600E in Google Scholar or PubMed, you'd need to search for every way that the variant can be described in text, such as p.V600E, p.Val600Glu, c.1798T>A, C1798A, etc. Furthermore, you'd need to then read the results to determine which are specifically referencing the V600E variant in the BRAF gene, versus mentioning the V600E variant in one of the other genes mentioned in the paper.
We've already done that work ahead of time, figuring out which variants belong to which genes in papers, normalizing all the different nomenclatures used for variants so that you can do a single search for your variant and remove the need to think about how authors could have referenced it.
We do the same with taking into account all the different ways authors can talk about genes, diseases, clinical contexts, meanings and interpretations, etc.
I don't know if that answers your question. I guess it depends on what you mean by knowledge graph, since there seem to be a few different ideas of what that entails.