Our new arXiv paper: We built a generative AI model of blood glucose levels from 10 million continuous glucose monitoring measurements of 11,000 people from the Human Phenotype Project
We show that our model can predict clinical parameters such as liver-related parameters, blood lipids, and sleep-related indices. The model can also predict onset of future health outcomes even 4 years in advance, as well as outcomes of randomized clinical trials
We also integrated dietary data into the models and show that the enhanced model can accurately generate blood glucose levels using only dietary intake data, simulate outcomes of dietary interventions, and predict individual responses to specific foods.
The model also performs well on 15 different external datasets from 5 different geographical regions and 6 different continuous glucose monitoring devices
Paper: https://t.co/KbBo3DLnsa
Data: https://t.co/l9V3k0IuzY
Great work led by @GLutsker, Gal Sapir, @nastya_godneva, @smadarshilo, @H_Rossman in collaboration with @SamochaBonetD and @nvidia researchers @GalChechik@MannorShie@eli_meirom
Are there universal laws driving complex systems dynamics across scales? What kind of general theory connects fluctuations in microbiomes, rainforests or economic and urban systems? Check this @PNASNews paper by ashish george & @Jp_odwyer@sfiscience https://t.co/1ZBSmzXZli
For some individuals, adding physical + heart activity can help explain glucose variance, but we only saw this for a subset of individuals. Thanks again to Tin-Hai and Felix for the help on this project, which has been a nice collaboration that started with our PHRT funding.(5/5)
Very happy that our article "Uncovering personalized glucose responses and circadian rhythms from multiple wearable biosensors with Bayesian dynamical modeling" is now online in Cell Reports Methods! (https://t.co/XUVZZ5SwFB) A brief overview... (1/5)
A few years ago we published a method on detecting oscillations in single-cell time series with @MagnusRattray @PapalopuluLab. We now have a basic implementation using Python and @GPflowProject ... Jupyter Notebook tutorial here -> https://t.co/jUCTVmiivz
How precise are noisy #morphogen gradients in #tissue#patterning?
It turns out that the positional information they convey to #cells in the #neuraltube is much more accurate than previously estimated!
Our new #devbio paper is out in @NatureComms: https://t.co/oIkk5R7LF9
🥳 Just awarded a Synapsis foundation grant !
➡️Open PhD student position
🔬🧠Protein homeostasis in live single human neurons in healthy/ND disease.
If you feel 🤩 tell me and apply at @epfl_edcb or EDMS PhD program. Thx for RT
I've written a tutorial on Bayesian inference for single-cell gene expression data using STAN @mcmc_stan. We see how the Poisson, negative binomial and beta-Poisson mixture distributions emerge from simple stochastic models of gene expression dynamics. https://t.co/Qq4lSImAwe
smFISH of circadian clock genes + mixture models shows that circadian time constitutes only a small fraction of the total variability in mRNA number between cells. Our new pre-print @Naefelix@jakeyeung
Our tightly collaborative work with @Naefelix and @NPSysBio is out - Transcriptional memory varies widely between genes and scales with expression variability - may play a role in tissue patterning
https://t.co/HuASPACHmo
Pleased to announce that I'll be starting a 2 year @PHRT_CH fellowship. We'll use Bayesian modelling, app data and transcriptomics to investigate the relationship between the time at which we eat, our circadian clock and metabolic disorders.