📢 Sydney #RNA scientists: our next #SydneyRNASalon event is coming up! 🧬 ECR, student, or just RNA-curious? This one's for you!
📅 Thu 30 April · 2:10–4:30pm
📍 Room G07, Biosciences Building D26 UNSW
🥐 Refreshments sponsored by @rnasociety.bsky.social, Lexogen & @ramacgenomics.bsky.social
I don't think protein structure is generally required for mRNA vaccines (but it is for docking experiments, e.g. drug binding simulations)
Neoantigen discovery requires genomic profiling (exome might be sufficient). These mutant proteins are what provides the immune system a cancer-specific target
Paul Conyngham AI entrepreneur and Rosie's owner, with Prof. Pall Thordarson (myself), Director of the UNSW RNA Institute plus Rosie the staffy getting ready for our interview on Channel 9 Today Weekend show last Saturday.
@ashjogalekar.bsky.social Summarised beautifully on X the most profound lesson re Rosie and her personalised canine cancer vaccine, namely that Paul Conyngham's work here using AI to design the mRNA we made, demonstrates that Dyson's prophecy about "domesticating" Biotech is within reach! #ozchem
AlphaFold was used for molecular docking experiments to identify small inhibitory molecules, but the identified drugs could not be bought/imported, so this led us down the mRNA/neoantigen path
I think it's really important to find trusted, open-minded experts in biomedical and clinical research to guide you should you embark on a similar journey.
Agree the Alpha fold part isn't well reported. This is a question for Paul. I believe he used it to model mutated protein structures, the used molecular docking simulations to find a small molecular inhibitor. Ended up hitting a brick wall with that, as reported, which led to mRNA
Paul has a blog post and the University has a media piece coming out as well.
We're focused on the science here, experiments underway for eventual publication.
It's a good question I don't think I can accurately answer, but I think there's potential for "democratization" of this technology, perhaps one of the best ways to make it scalable for humans.
This might interest you:
www.scientificamerican.com/article/pers...
Probably correct, but we're comparing apples to oranges here. Don't think dogs get the same level of care as most humans, or at least don't have as many therapeutic options.
It was far from inexpensive!
Another question for Paul, but I believe yes is the answer. Probably with some AI-influenced filtering beforehand
Paul performed these steps with the help of ChatGPT. The resulting pipeline isn't public... Yet
This is a question for Paul, but I suspect this is unlikely as there may be sensitive intellectual property embedded within the pipeline.
You could use ChatGPT to come up with something else though, like Paul did!
It's less than 2% of your genome ;)
But yes, sequencing costs have gone down... For consumables. Skilled staff wages and quality control / data analysis are still major components of a sequencing service cost recovery model. Throughout helps bring cost down
Yes, we do this regularly in the research setting. What's remarkable here is that Paul was able to get upskilled and understand the biology with ChatGPT. It helps that he could afford experiments and had data analysis / AI experience (like how to launch high performance computing jobs)
AU$3k was the initial tumour and normal DNA sequencing data, generated in early 2025. That cost was due to a balance of speed and throughout.
We performed additional sequencing thereafter, to help validate the neoantigen selection, and now to see what's happening in the treatment-unresponsive tumor
Need any further clarifications?
A good news story for you this morning in The Australian, featuring #genomics, #AI, and #mRNA #vaccines 🧬🐕
www.theaustralian.com.au/business/tec...
The deadline has been extended to March 15th for the RNA Society's Infographics contest! Get your talent team together and submit! Prizes!! Go to www.rnasociety.org/rna-society-... or scan the QR for details!
Could @lornegenome.bsky.social be the MOST Aussie genomics conference?
I reckon there could easily be 3 versions of this, there's already two...
Zheng Su presents kiloBrains.com, an AI-powered biodata analysis platform.
A virtual bioinformatics assistant
#AGTA2025
Cornetto from @hasindu2008.bsky.social presented by Ira Deveson: T2T assembly of human genomes with @nanoporetech.com adaptive sampling by rejecting the "boring bits". 27/46 chromosomes fully assembled from one flowcell. Can also be used on saliva!
#AGTA2025
github.com/hasindu2008/...
Check out his team's online ressources for long read songle-cell sequencing
kinnex-documentation-external.readthedocs.io/en/latest/in...
And keep an eye out for LRAA...
"Isoforms are important; how come we're not studying them?"
#AGTA2025 keynote @azizalkhafaji.bsky.social mentions that we are now at am I flection point, where 100s of millions of reads are produced per run.
The great Prof Vanessa Hayes, who sequenced the first African Genome, states "the closer we get to ancestral populations, the more diversity we see in genomes".
Presents genomic data from WGS of Khoe-San populations, interesting phylogenetic split between Khoe (hunters) and San (gatherers) #agta2025
Prof Sean Grimmong from PeterMac Genomic Cancer Medicine on the entire workflow for #cancer #diagnosis, and how to deal with dilemmas, like the 5% of Australians without a clear diagnosis #agta2025
