🔥Telegram İfşa

What is life as a PhD student like? 🔬 Current PhD candidate Felix Kohane (Felix Kohane) from the SystemsMicrosopy_Australia lab at UNSW and Cancer Cell Plasticity Lab lab at Garvan Institute of Medical Research shares his insights ⬇️

"I was initially attracted to biomedical sciences through my fascination with the cellular origins of disease. This led me to the field of cancer research, where I developed an interest in the use of imaging techniques to capture and model cellular behaviour."

"A typical day during my PhD involves a bike ride through Centennial Park on the way into the lab. Once there, I make sure my cell cultures are happy and healthy before focussing on either planning, conducting, or analysing my experiments."

Got high-dimensional #singlecell #omics or #imaging data & considering how to prepare, perform and analyse in a lower dimensional setting? Check our perspective on #RepresentationLearning Cell Reports methods with Fatemeh Vafaee t.author.email.elsevier.com/r/?id=h585f3be…

This is the first paper from #PhD candidate #IhuanGunawan, and another from our collaboration linking UNSW Medicine & Health UNSW Engineering UNSW Science UNSW AI Institute @uDASH_UNSW with support from #ScientiaFellowship Evelyn Mike #UNSWScientiaFellows

Our new collaborative perspective on #BiomarkerDiscovery for #PrecisionMedicine in advanced #ProstateCancer. This work nucleates a new Australian national network dedicated to improving the treatment journey for PCa patients. link.springer.com/article/10.100…

Come join us in Sydney, bring your passion for #research, your drive to fight #cancer, and your love of all things #imaging, data and teamwork. UNSW @uDASH_UNSW

Oh, hey! My talk video is online if you'd like a tutorial on AI in preclinical drug discovery: "Phenomics in Drug Discovery: Microscopy and Machine Learning" from the 2024 Machine Learning for Drug Discovery Summer School hosted by Mila/Valence Labs youtube.com/watch?v=mc3vVx…

Using #fluorescenceimaging in your research? Wish you could quantitatively analyse 10s of markers per cell, while using the 4-plex methods you are used to? Our Extensible Immunofluorescence (#ExIF) framework lets you do it! Check out our preprint here: researchsquare.com/article/rs-487…

Using concepts from multi-modal #Omics data integration, Extensible Immunofluorescence (ExIF; shorturl.at/d3N75) allows highly multiplexed single cell analyses of complex biology (here of #EMT) by integrating standard 4-plex #fluorescenceimaging data via virtual labelling.

ExIF yields high-fideltiy, high-plexity image data integration by combining an extensible labelling strategy (recurring 'anchoring channels' + unique 'variable channels' in each 4-plex panel) with optimised virtual labelling via residual vision tranformers #generativedeeplearning

@RArgelaguet Excitingly, we see that ExIF dramatically enhances downstream quantitative single-cell analyses of complex EMT biology, firstly by hugely improving machine learning-based classification (i.e. recognition) of experimentally perturbed EM-state phenotypes.

@RArgelaguet We also show that ExIF massively improves the structure of t-SNE, UMAP & PHATE manifolds based on quantitative single cell feature measurement using #CellProfiler (thanks to Anne Carpenter, PhD for the gift that keeps on giving).

@RArgelaguet Anne Carpenter, PhD Krishnaswamy Lab Igor Adameyko Psuedotime analyses in PHATE manifolds based on ExIF integration show epithelial & mesenchymal marker dynamics matching established patterns of repression and enhancement during EMT. Again, these dynamics cannot be inferred from standard, non-integrated immunofluorescence data

@RArgelaguet Anne Carpenter, PhD Krishnaswamy Lab Igor Adameyko Thanks to #IhuanGunawan for leading this work with support from #ErikMeijering, Felix Kohane, Fatemeh Vafaee, UNSW, UNSW Medicine & Health, UNSWEngineering, UNSW Science, UNSW AI Institute, Evelyn Mike, #UNSWScientiaFellows

In #cellbiology, we're used to the idea of information flowing by A binding to B. But with finite numbers of A, the AB interaction reduces A's other interactions. How well do we map such negative information effects & their network propagation using epistatic perturbations?

For creatively thinking about & tackling really #complexproblems, which of the following do you think achieves the best process & outcomes? Why? What is required in each situation to optimise outcomes?

When the Financial Times puts #OpenAccess in air quotes, we researchers who would prefer to spend public funds on research rather than private publisher's fees should realise that "Open Access" is the wrong solution to the right problem. We can do much better. ft.com/content/e91a96…

For those looking to invest in the #SpringerNatureIPO noted by Financial Times, please note the class action lawsuit filed against them and 5 other large academic journal publishers for contravention of Sherman Act anti-trust laws. lieffcabraser.com/antitrust/acad…

Our #AI #research Nature Communications will help you use concepts from #multiomics data integration to fuse 4-channel imaging data into unlimitedly multiplexed image datasets to interrogate complex cell biology, like #EMT. rdcu.be/emtBc UNSW UNSW Medicine & Health Fatemeh Vafaee