I’m sending this to my medical oncologist at Anschutz ( I’m a phase 2 trials patient for Ippi/nivo). We just talked about the coalescing of large data sets for research/ doctors to use in patient care.
Quick question on understanding the importance of delta from your chart: when comparing two different biomarker > targets, the one with the "higher" negative delta number means that specific biomarker to target has a better (?) synthetic lethal interaction with cancer cells than another? And how does the Adjusted r specifically relate to the delta number? Trying to understand your chart so I can better understand the context behind your Codex experience, thank you!
Hey, yes! A bigger delta is a better synthetic lethal interaction (kills more cancer cells than normal cells).
The r is the correlation between alteration status and gene effect. So, for example, DNMT3A WT cells get assigned 0, and mutant cells get assigned 1. As DNMT3A status moves from 0 to 1 across different cell lines, the effect of ASF1B loss becomes more negative (cells are more dependent on ASF1B). The adjusted r is this same number, after removing effects of confounders that might explain the correlation (like cancer tissue type, mutation burden, microsatellite instability status).
A) I think this pretty conclusively demonstrates that AI can discover new things and genuinely do science.
B) Samuel is right that we're close to enabling single person biotech companies that can make and test discoveries in days or weeks which previously required large teams working for years.
I’m sending this to my medical oncologist at Anschutz ( I’m a phase 2 trials patient for Ippi/nivo). We just talked about the coalescing of large data sets for research/ doctors to use in patient care.
Btw: my medical oncologist was at ASCO with you Chicago 2026
Quick question on understanding the importance of delta from your chart: when comparing two different biomarker > targets, the one with the "higher" negative delta number means that specific biomarker to target has a better (?) synthetic lethal interaction with cancer cells than another? And how does the Adjusted r specifically relate to the delta number? Trying to understand your chart so I can better understand the context behind your Codex experience, thank you!
Hey, yes! A bigger delta is a better synthetic lethal interaction (kills more cancer cells than normal cells).
The r is the correlation between alteration status and gene effect. So, for example, DNMT3A WT cells get assigned 0, and mutant cells get assigned 1. As DNMT3A status moves from 0 to 1 across different cell lines, the effect of ASF1B loss becomes more negative (cells are more dependent on ASF1B). The adjusted r is this same number, after removing effects of confounders that might explain the correlation (like cancer tissue type, mutation burden, microsatellite instability status).
nice...
A) I think this pretty conclusively demonstrates that AI can discover new things and genuinely do science.
B) Samuel is right that we're close to enabling single person biotech companies that can make and test discoveries in days or weeks which previously required large teams working for years.
Can AI at least summarize and appraise the literature as well as a good traditional Systematic Review?