Daniel Temko

Delighted to share a substantially updated preprint led by @gladyspoon1 where we used single-cell derived phylogenetic trees from 80 AML patients to show that pre-leukamic evolution is driven by pervasive positive selection biorxiv.org/content/10.110…

@SamPowerful1 Hard yes

This guy asks if he can watch the solar eclipse with you at your house. What is your answer?

@sp_monte_carlo Hmm… yes if one is OK with hypotheticals, I agree. But there’s the rub. What’s a hypothetical. A mysterious possible world? Maybe twitter’s not the right forum for this.

@DanielTemko Not much to say! The maths checks out, and it can be used as a basis for active decision-making, so perhaps there's a sensible interpretation somewhere among those. It's all pretty manageable if one is comfortable with hypotheticals, I guess.

So what is probability anyway? Does anyone have an answer satisfied with? Put another way: what does it mean to say there is a 50 percent chance of a coin coming up heads?

@nomad421 UMAP has been an integral part of many discoveries from scRNA-seq. It’s a great tool for revealing non-linear patterns at low computational cost. If there was no room for improvement we’d all be out of a job. But not useless!

During routine EDA with some single-cell RNA seq data (#scrna), a UMAP plot reveals cluster structure that later turns out, through orthogonal & rigorous validation, to be real and meaningful. Can/should one put the figure that suggested that subsequent analysis in the paper?

@littmath I spent more hours of my life than I care to admit thinking about a variant of this problem. The answer is "red", and the key to seeing this is to note that the colours of the balls are correlated. So, if you know one ball is red, others are more likely to be red than green 1/2.

You have 100 urns; each contains 99 balls. In 99 of the urns, one ball is red and the rest are green; in the last, all 99 balls are red. You choose one of the urns uniformly at random, draw a ball—it’s red—and throw it away. The next ball you draw from the SAME urn is probably:

@littmath I spent more hours than I care to admit thinking about a variant of this problem. The key is to notice that the colours of the balls are correlated, and so if you know one ball is red, others are more likely to be red than green. 1/2

You are given an urn containing 100 balls; n of them are red, and 100-n are green, where n is chosen uniformly at random in [0, 100]. You take a random ball out of the urn—it’s red—and discard it. The next ball you pick (out of the 99 remaining) is:

Check out this comprehensive, large-scale, multiplatform, multiomics study recently published in @CellReports on new heterogeneity and transcriptional drivers of #TripleNegativeBreastCancer. #TNBC #BreastCancer @LabPolyak @DanielTemko @ANewShree @nellage #MylesBrown

@nmancuso_ Can you explain what the problem is you’re referring to? A bit odd to specify the distribution of X, but otherwise this looks links a standard linear regression.

am i being pranked rn

@ItaiYanai @MartinJLercher Excellent. But is any analysis ever truly hypothesis-free? Maybe a better term would be "hypothesis-light"?

12. And remember, always plot your data! There could be gorillas hiding in there! genomebiology.biomedcentral.com/articles/10.11… @MartinJLercher

Here are my 12 guidelines for data exploration and analysis with the right attitude for discovery: 1. You never really finish analyzing a dataset. You just decide to stop and move on at some point, leaving some things undiscovered. 🧵

How accurate is #histone #inheritance? @Architex_97 & @shaon_chak propose a new model which suggests that the accuracy might be lesser than currently believed. The pre-print of the study is now available on @biorxivpreprint: biorxiv.org/content/10.110…

New preprint on mutation accumulation in growing populations with David Cheek and Tibor Antal: arxiv.org/abs/2208.02088 1/n

@yosoykit @trevoragraham @samirguptaGI @AmerGastroAssn Congratulations Kit!

Very excited to share that I am a 2022 recipient of the AGA Research Scholar Award! Lots of work in IBD and mathematical modeling to come, in great AGA company. Special thanks to mentors @trevoragraham @samirguptaGI for all the support. #IBD #mathonco thank you @AmerGastroAssn

Noted and will incorporate into 3rd edition @HastieTrevor @robtibshirani @gareth_m_james

My Computational Radiation Biology and Oncology Lab @RadNetCRUKCoL @UCLmedphys @uclcancer will soon advertise for a postdoc to work on mathematical/statistical modelling to optimise drug-radiation combination therapy schedules. DM me to discuss. Please RT. #mathonco #radonc

Correlation Vs. Causation

We developed an experiment to increase the scale of QTL mapping so that we may observe how 'low effect' variants play a role in determining phenotypes. 1/4

@BenjaminWerner_ @QMBCI Congratulations Ben!

I feel so honoured to be listed alongside such great scientists to receive a UKRI Future Leaders Award #UKRIFLF to continue (and extend) my work on the evolutionary dynamics of extra-chromosomal DNA @QMBCI.

@FryRsquared A tangy delight

Settle an argument for me? Pineapple on a pizza is: