←
Mile Sikic1,892 posts
Mile Sikic avatar

Mile Sikic

@msikic

Group leader and assistant director @astar_gis, Prof @SveucilisteZG, AI in Genomics

◷ Joined December 2008▧ 36 Photos and videos
Mile Sikic avatar
Mile Sikic@msikic·Apr 28

HERRO has been published in @Nature !! nature.com/articles/s41586-0… This achievement is a result of the great work by @domstanojevic, with contributions from @DehuiLin, @sergeynurk, and @PaolaFlorezdeS We are entering the era of high-quality genome assemblies supported by AI. This is just the beginning.

Sergey Ovchinnikov avatar
⟳ Mile Sikic retweeted
Sergey Ovchinnikov@sokrypton·Nov 14

"Since the advent of AlphaFold, homology search has been an integral part of structure prediction" What do you mean? MSA has been an integral part of structure prediction since 1990s. 🤪

Hannah Wayment-Steele avatar
⟳ Mile Sikic retweeted
Hannah Wayment-Steele@HWaymentSteele·Nov 13

Really excited to share! In this day and age the hype might suggest we’ve “solved” protein biophysics, but all the models we tested are far from predicting simple N chemical shifts. Hoping releasing our cleaned / curated data and models can help all protein ML space better learn biologically-relevant chemistry and dynamics!

1/ 🧵 New preprint! Do protein foundation models really understand local chemistry? We probed them with NMR chemical shifts, per-atom reporters of chemical environment. Short answer: they struggle exactly where the biology gets interesting 👇 [link] biorxiv.org/content/10.64898…

Post attachment
Jim Shaw avatar
⟳ Mile Sikic retweeted
Jim Shaw@jim_elevator·Nov 8

Update: I will be starting in Fall 2027 as an Assistant Professor at UNC Chapel Hill in the CS department. My lab will build algorithms for large-scale microbial and meta-omics sequence analysis. I'll be recruiting PhD students. More info at jshaw-lab-compbio.github.io 1/2

Andrej Karpathy avatar
⟳ Mile Sikic retweeted
Andrej Karpathy@karpathy·Oct 2

We'll be spending a lot more time trying to understand the outputs of language models. A few thoughts, tips & tricks: Writing. Something I've had success with: Ask your LLM to explain something in ASD-STE100, it's a controlled language specification originally developed for aerospace maintenance documentation. LLMs well-versed in this language and it comes with heavy constraints on clean writing style that I often find a lot more readable. Sometimes I've tried to soften it a bit e.g. ask for "80% of the way to ASD-STE100" because the spec is quite stringent. But even better: Diagrams / images. Instead of writing, ask your LLM to create a diagram. These can be a lot easier to process, parse, and understand. But even better: Web pages. Ask for output "in HTML" to get a beautiful, interactive webpage. LLMs are getting really good at frontend and can create beautiful experiences, animations, etc. But even better: Explainer videos. The output format I am most bullish on is fully custom / bespoke explainer videos generated on any arbitrary topic. Experiment with things like "Create a 3b1b style video explainer on X. Use my ElevenLabs API key for audio narration". (you'd need an API key for the latter or you can ask your LLM to find you decent free alternatives that use your local compute). This is actually starting to work! In summary: - As LLMs get better, they will do more and more of the legwork autonomously, and a lot more of our work will rise up the abstractions into oversight and understanding. - Luckily, LLMs can help here too because as intelligence and code are increasingly abundant, you can ask for large, custom, discardable software artifacts (e.g. web apps, video explainers) that would have never made sense to create before. Push the boundaries here and you'll be surprised.

Post attachment
Thomas Norman avatar
⟳ Mile Sikic retweeted
Thomas Norman@thenormanlab·Oct 13

1/ Most genes and variants matter somewhere but look silent in any one cell model. Can we build conditions that reveal phenotypes? With @kanishkadey, we engineer cell states to expose hidden disease biology and trace how genetic risk reaches disease genes. biorxiv.org/content/10.64898…

Post attachment
Anshul Kundaje avatar
⟳ Mile Sikic retweeted
Anshul Kundaje@anshulkundaje·Oct 5

Just a few quick things. 1. Nice rigorous study of the importance of being careful with performance metrics and taking into account data issues. 2. This should be done BEFOFE one fits big models & make lots of big claims. Nevertheless better late than never. 2/

The "Single-cell foundation models don't work for perturbation prediction" narrative was a measurement problem. Our new paper is out in @NatureBiotech : "Deep Learning Perturbation Models Can Outperform Baselines on Calibrated Metrics" Last year, several papers concluded that deep learning perturbation models couldn't beat simple baselines, a finding that cast doubt on the entire Virtual Cell direction. We asked a different question: can the metrics used in those comparisons reliably tell a good prediction from a bad one? Often, they can't. Common metrics like MSE and control-referenced Pearson Δ frequently rank uninformative baselines above positive controls. With calibrated alternatives (eg, WMSE, weighted R²Δ, NIR), deep learning models do outperform mean, control, and linear baselines. The data matters too. Norman19 is approximately 96% additive, and models see less than 1% of possible gene combinations during training. The ceiling for beating an additive baseline is nearly zero, and the training signal for non-additive interactions is almost absent. Some dataset rankings aren't model rankings, but just noise! To address this, we developed the Dynamic Range Fraction (DRF): a measure of metric calibration. Before trusting a model ranking, verify that your benchmark can detect a useful prediction at all. The models aren't shiny toys. The field just needed better rulers. Paper: nature.com/articles/s41587-0… Code: github.com/shiftbioscience/P… Huge congratulations to the first authors @Henrymiller2012 and Gabriel Mejia, and the team of Shift Bio!

Post attachmentPost attachmentPost attachment
Mile Sikic avatar
Mile Sikic@msikic·Sep 9

We need more projects like apertus-ai.org/ which come from academia. Open source and open data are more valuable than ever

Don't be like these guys. Don't try to secretly scoop others when u hear about their potential success especially those with far less resources & power. Learn to give credit where it is due. Collaboration & healthy competition is the best way forward for science. Not this.

Roma Nagle avatar
⟳ Mile Sikic retweeted
Roma Nagle@nagle_roma·Sep 4

RNASSTR is out! Check out this article on our RNA secondary structure dataset, which leverages Rfam to build a larger corpus with structure-aware train/test splits for RNA secondary structure prediction models. innovativegenomics.org/news/…

Anshul Kundaje avatar
⟳ Mile Sikic retweeted
Anshul Kundaje@anshulkundaje·Sep 1

I am really not sure what Anthropic is trying to achieve here. Showing us great benchmark scores for Fable 5/5.1 for bio applications in their internal evals but making the model essentially unusable. Absolutely ridiculous strategy. Not even a trusted users option?!?

I think Anthropic’s safety guardrails are becoming a joke to biologists. Asking “what is a protein?” definitely doesn’t need extreme bio-safety restrictions, but “no need” and “not possible” are very different things.

Post attachment
Lior Pachter avatar
⟳ Mile Sikic retweeted
Lior Pachter@lpachter·Aug 29

An underappreciated impact of AI on biology is that an enormous, I mean ENORMOUS, number of skeletons buried in genomics papers are all about to burst out of their Supplementary Materials.

Post attachment
Daniel Liu avatar
⟳ Mile Sikic retweeted
Daniel Liu@daniel_c0deb0t·Aug 28

If you are working on an important bioinformatics tool, I encourage you to apply to our AI for science program for Claude credits. Link in 🧵

Shyam Prabhakar Lab avatar
⟳ Mile Sikic retweeted
Shyam Prabhakar Lab@shyam_lab·Aug 26

Back when my lab did a lot of ChIP-seq, we spent a lot of time validating our antibodies. Roughly half were bad. Now we know why.

"More than 18,000 questionable images found in antibody catalogues of 15 companies" - It's stuff like this that makes bio research so slow

0:06
Mile Sikic avatar
Mile Sikic@msikic·Aug 18

Can we cure most human diseases in the next 5–10 years? Absolutely not. Is it still the right thing to aim at and invest in? Absolutely yes

2/2 Second, on the messaging around AI.  I do not agree that my messaging has been disproportionately negative.  In fact it has been about equally balanced between risks and benefits: I’ve written one major essay about each, and even in interviews where I discuss the risks, I make sure to frequently mention the incredible benefits as well as proposing possible solutions to the risks (short clips from my interviews that end up on social media tend to be disproportionately negative, as that gets clicks).  In fact, I wrote Machines of Loving Grace because I didn’t feel the AI industry was painting an inspiring enough picture of how the technology could radically transform the world for the better.  The bulk of the essay is devoted to refuting skepticism of AI’s potential in health and biology, and showing why I think it will actually be possible to cure most human disease in ~5-10 years, as crazy as it may sound to ordinary people and frankly to biologists as well (I used to be one!).  And, if you read my most recent essay (Policy on the AI Exponential), I discuss concrete proposals for how to streamline the FDA process to make sure the deluge of AI-accelerated drugs isn’t slowed down by the regulatory process.  I feel the urgency here: I lost my father to Hepatitis C only a few years before the development of direct-acting antivirals (sofosbuvir), which cure 95% of patients and probably would have cured him. I do agree that the public has a negative view of AI (and that this is a big problem), but I don’t think it is primarily caused by me or any other AI leader warning about AI’s risks.  I think it is fundamentally a crisis of trust.  I think that ordinary people don’t trust companies, governments, or the tech industry and always suspect that we are cooking up some new way to screw them over.  The causes of this go back decades and AI is just the latest iteration of it.  I don’t think that a glitzy marketing campaign with a positive spin (which some have advocated that Anthropic do) is the way to win back that trust — at this point, saying that AI will cure cancer is more a cliche than it is inspiring, and most people think it is deceptive.  The thing that will work is *actually curing cancer*.  I think by far the most accurate criticism of AI companies including Anthropic is that we haven’t yet delivered on our big promises to benefit the world.  That is totally on us, and I think it’s the criticism you should be making, instead of all this stuff about messaging and marketing. We are however doing our best to fix this: Anthropic is ramping up its efforts very quickly in biology and medicine, and we hope to have incredible results in the coming years and some early glimmers in the coming months.  When we’ve actually accomplished something real, the whole world will hear about it, as loudly as possible, you have my word on that.  But until then I don’t want to make empty promises, and in the meantime I feel compelled to speak honestly about the very real risks of AI and how to address them.  Honesty is the right thing on the merits, and in terms of public credibility and trust it is no worse than, and may in fact be better than, an approach that ignores or distracts from risks which people instinctively understand are real.