Just out - the rise and fall of SARS-CoV-2 lineages in England. In the last 1.5yrs the UK has been a bell weather for SARS-CoV-2 evolution and genomic epidemiology thanks to the data sequenced by @CovidGenomicsUK and @sangerinstitute. Let's recap. >>

This summer we'll be opening a new lab at the German Cancer Research Centre @DKFZ in Heidelberg. Come join us studying cancer evolution with spatial & single cell genomics + AI. There will be plenty of openings for students, postdocs, and technicians; dry and wet lab. >>

In this paper we describe the multifaceted mutation patterns observed in 61 single, double and triple knockouts of DNA repair genes in C. elegans. Phenomena range from point mutations, over complex rearrangements to local hypermutation.

As part of the #PanCancer #PCAWG project we studied the evolutionary history of 2,658 cancers. Here's what we found: Whole genome duplications and chromosomal instability occur several years to decades prior to diagnosis 1/6 https://t.co/DPZ7ilwlOH

Excited to share our latest @GenomeBiology paper, which came out in early January "Clustered CTCF binding is an evolutionary mechanism to maintain topologically associating domains"! https://t.co/jp2DDdE0Q8

@naturemethods @metricausa @jccaicedo @broadinstitute Like the classic #ImageNet competitions that accelerated the era of deep learning, the Human Protein Atlas competition had 2,171 team compete for analyzing ~78,000 images--> high performance models https://t.co/GiYsucJsa9 @Prof_Lundberg @weioyang @KTHuniversity and collaborators

Or it identifies a signature of somatic hypermutation producing clustered mutations with a distinct mutation spectrum at transcription start sites in lymphoid neoplasms. This appears to be the signature of AID and is different from the genome-wide clustered mutation spectrum 4/5

Want to learn mutational signatures jointly based on mutation types + genomic activity patterns? My student @harald_voeh has developed TensorSignatures to better characterise and localise mutational processes. https://t.co/ILXgBCJCpZ 1/5

My EMBL 2nd year PhD friends/fellows are doing a great job organizing this years' PhD Symposium, check it out! ;)

For those at #EMBLCanGen. Check out @harald_voeh’s poster 147 today. How to learn mutational signatures from mutation spectra *and* genomic activity patterns with TensorSignatures.

It also finds a lot of associations in bulk transcriptome data, deconvolves the signal to find areas on each slide corresponding to molecular cell types such as tumour infiltrating lymphocytes. Entirely automated. 3/5

Similarly the network finds prognostic associations in most cancer types that match and augment conventional grading and subtyping and points out prognostically relevant regions, such as necrosis, on each slide. 4/5

This demonstrates how deep learning can integrate histopathology and genomics with huge potential for digitally augmenting diagnostic workflows - inc. challenges to tame a CNN. Great work by many lab members, Alex Jung and @luiza_moore. Facilitated by #TCGA and #METABRIC. 5/5 ///

Hello world. Here’s something interesting: @yufu0413 from my lab trained a deep convolutional neural net in cancer histopathology *and* genomics using 14M images from 17k H&E slides across 28 cancer types. The outcome is stunning. 1/5 https://t.co/H3lGmoSKn8

Without spatial annotation of tumor lymphocyte, our method PC-CHiP (Pan-Cancer Computational HistoPathology) is able to automatically identify regions with lymphocytes for thousands of large H&E whole tissue slides. Great tool to assist pathologists!