Returning the favour: @ara_anderson closes the MathMed conference talking about the power of adaptive therapy with a 2001-inspired intro!

A week left before the deadline! ⏰ Come join our team at @StAnna_CCRI!. #Technician #Cancer #Evolution #Resistance #Pediatric #Genomics

Sounds like an amazing opportunity!.

What a great event! Thank you so much for the opportunity to be a part of it!.

Thank you for hosting and inviting me! I very much enjoyed a great day chatting about the overlaps between popgen and cancer evolution!.

Cool lab! Cool work!.

Fantastic opportunity!! Come join us!.

RT @cancerbits: My lab @StAnna_CCRI is recruiting for a new #Postdoc scientist. This time we are looking for a genomics data analysis wizar….

RT @DSeruggia: First pre-print from the lab! Together with @lucapinello and @danielevanbauer we developed CRISPR-CLEAR, an experimental an….

An excellent opportunity with a great scientist and person. 🙂.

RT @PKameneva: PhD position alert 🚨 I am looking for 2 PhD students to join my group on cancer initiation at @StAnna_CCRI. #FWF funded proj….

11. Huge thanks to all co-authors! It was a real team effort! Special thanks to Kat Webb for PCa expertise and a big thank you to @AndreaSottoriva and David Dearnaley for conceptualising and guiding the project. Cancer evolution isn’t just cool, it’s also meaningful for patients!.

10. Strikingly, only patients that were both genomically and morphologically heterogeneous had notably shorter time to recurrence. Indicating that combining measures is needed for full benefit! We propose that diversity, evolution's fuel, predicts outcome at diagnosis in PCa.

9. Independently, genomic and image analysis produced metrics that were significant in multivariate analysis, where (importantly) we controlled for common clinical co-variates. As Darwin recognised, variation is key to evolution, and both metrics were related to diversity!

8. Ultimately, we wanted to see if evolution metrics were meaningful for patients. Prior to unblinding ourselves to patient outcomes, we settled on a series of metrics to quantify genomic instability, heterogeneity, and morphological diversity. This led to lots of brainstorming!

7. We also measured mixing of cancer and immune cells. Chromosome 6p loss correlated with reduced mixing of cancer and immune cells, implicating a usual suspect, HLA LOH, in immune evasion. This study shows the power of combining image analysis and genomics to understand biology.

6. Our AI-derived segmentation allowed us to average the area covered by different Gleason patterns, a metric we termed "Continuous Gleason" that captured subtle differences in tissue grade. This gave us more power to reveal the association of dedifferentiation and aneuploidy.

5. To power up our analysis of tissue morphology we used AI to Gleason grade individual glands. We used this to produce a standard Gleason score, and to measure the diversity of Gleason patterns. We also identified epithelial, stromal and immune cells using an AI cell classifier.

4. The IMRT trial (Royal Marsden) studied toxicity in locally advanced PCa and was key to our study. Diagnostic 12-needle biopsies from IMRT patients provided multi-sampled tumour, and we assessed genetic profiles alongside matched tissue slides, linking genotype and phenotype.

3. We know that cancer's ability to evolve drives recurrence, but can it be used to make predictions? In this paper we measure PCa "evolvability" at diagnosis and study the relationship between mutations and phenotype to, ultimately, forecast patient outcome up to a decade later.