Peter
@biotechpedro.bsky.social
PhD student at Fraticelli lab @irbbarcelona.org
I try to convert coffee into code and ideas for uncovering biological mysteries.
Some interests: single-cell, lineage tracing, computational biology, cellular variability, premalignancy, resistance...
I try to convert coffee into code and ideas for uncovering biological mysteries.
Some interests: single-cell, lineage tracing, computational biology, cellular variability, premalignancy, resistance...
Reposted by Peter
Training other methods on the same data may not be feasible due to compute limitations. Training on other much smaller data won’t be the same task. Training your implementation or with your compute limitations is not representative of the original method.
October 26, 2025 at 10:51 PM
Training other methods on the same data may not be feasible due to compute limitations. Training on other much smaller data won’t be the same task. Training your implementation or with your compute limitations is not representative of the original method.
Reposted by Peter
Academics also can be driven and intelligent, but the driver is often personal curiosity and understanding, with a fair bit of showing off (academic egos are something else; I am self aware enough to realise I am not immune to this though I do try to tame it).
October 26, 2025 at 7:48 AM
Academics also can be driven and intelligent, but the driver is often personal curiosity and understanding, with a fair bit of showing off (academic egos are something else; I am self aware enough to realise I am not immune to this though I do try to tame it).
Reposted by Peter
Are professors who don’t teach or mentor, just academic executives who raise money and market other people’s work? When did teaching stop being an important part of this job?
October 25, 2025 at 5:26 AM
Are professors who don’t teach or mentor, just academic executives who raise money and market other people’s work? When did teaching stop being an important part of this job?
Also, docs.omnibenchmark.org/latest/ by @markrobinsonca.bsky.social
doi.org/10.48550/arX...
doi.org/10.48550/arX...
doi.org/10.48550/arX...
doi.org/10.48550/arX...
Omnibenchmark (alpha) for continuous and open benchmarking in bioinformatics
Benchmarking in bioinformatics is a process of designing, running and disseminating rigorous performance evaluations of methods (software). Benchmarking systems facilitate the benchmarking process by ...
doi.org
October 24, 2025 at 11:48 AM
Benchmarking should be a continuous effort by some sort of stable community (e.g. scverse), in a thorough and independent way, e.g. dreamchallenges.org
October 24, 2025 at 9:23 AM
Benchmarking should be a continuous effort by some sort of stable community (e.g. scverse), in a thorough and independent way, e.g. dreamchallenges.org
It's a VERY important point. Rather than benchmarking within a paper (which obviously is not independent btw), I'd prefer comp method papers to be (i) detailed in technical explanations, (ii) easily and fully reproducible, with multiple case studies. 1/2
October 24, 2025 at 9:23 AM
It's a VERY important point. Rather than benchmarking within a paper (which obviously is not independent btw), I'd prefer comp method papers to be (i) detailed in technical explanations, (ii) easily and fully reproducible, with multiple case studies. 1/2