Mark Derbyshire 🐑

Overhead irrigation very satisfying to watch.

Big reveal today! After uncovering the #sclerotinia in canola trial at @curtinuniversity.bsky.social for the first time in a week, we see exceptionally good and seemingly consistent infections! @theccdm.bsky.social @tobyenewman.bsky.social

Saving the field trial version 2.0. First a severe storm, now an early heatwave. Not only are the plants not overheating under the covers, evaporative cooling seems to be keeping it a few degrees cooler than outside. So far, some nice #sclerotinia lesions under there, which is what we like to see!

Been on a journey recently. Chasing consistent #sclerotinia across 660 randomised canola plants in a field plot led to temporary hoop-houses for uniform misting. Worked for two days then damaged by the storm. Managed to improvise a fix better than the original. Very grateful for my wonderful team.

Loving the #2025ISMPMI session on cross-kingdom movement of RNAs - so many different types of RNAs on the move. So much to learn!

Grateful for the opportunity to discuss our research. We've come a long way toward establishing the basics for improving sclerotinia resistance in canola. Ultimately, there's still a long way to go but many opportunities on the horizon with new methods like speed breeding and genomic prediction.

Well, I think most of them. Some issues with poor penetrance in dense canopies but generally the mainstay of protection. Purely anecdotally, if you just waft a fludioxonil plate by a sclerotium, it seems to work. Any thoughts?

📣 We have a postdoc position in fungus-microbiota interactions! Apply here: jobportal.uni-koeln.de/bewerben/2197
⁩
For more details, visit: bewerbungsmanagement.uni-koeln.de/ausschreibun...

Scientists with ADHD speak up: when fire meets focus Researchers with ADHD discuss its peaks and valleys and how they structure their work lives to succeed.

Less practically useful than I thought it might be but turned out to be a nice read. Encouraging to see that others have their own diverse ways of dealing with their ADHD, including playing to their strengths. Upvote mental health awareness in academia, for everyone.

doi.org/10.1038/d415...

🌿 A big week for CCDM with staff & students showcasing their research at two major conferences.
📍 #BotryScleroMoni2025 (Greece): Sharing insights on Sclerotinia stem rot and Botrytis cinerea.
📍 #APPS2025 (Sydney): Presenting work on fungicide resistance, cereal tolerance & disease control. 🌾👏

Heading home from #Botryscleromoni 2025. All over so fast! Feeling inspired and ready to start exploring new ideas. Thanks to all involved for a fantastic conference!

Well, either way it's clearly the 'best chemical agent'...

Good point.

#Botryscleromoni day 2. Another great talk by Daohong Jiang, thinking outside the box. Non-infective sclerotinia strain colonised with a mycovirus as a control for... sclerotinia. May be as effective as fungicides.

#Botryscleromoni day 2. Wonderful inspiring talk from Xin Li. Two nuclei = one haploid genome in #Science (www.science.org/doi/10.1126/...). Nucleus besties stick together through thick and thin.

In the midst of all the crazy things happening in the world, it has been nice to have a morning stroll through sunny Melbourne. Good for the soul. Feeling lucky.

It was a pleasure presenting today. Had some really enjoyable and insightful discussions at day one of #GRDCUpdates, and saw many great talks. Looking forward to tomorrow.

Really looking forward to this event. Always great to see all the exciting grains R&D going on across the industry.

Something oddly satisfying about learning how large language models work from Google Gemini. It's also weirdly accurate on this topic. Lots of training data online and relatively simple algebra... or emergent narcissism?

Genomic Foundationless Models: Pretraining Does Not Promise Performance The success of Large Language Models has inspired the development of Genomic Foundation Models (GFMs) through similar pretraining techniques. However, the relationship between pretraining performance ...

This is an important and interesting preprint: doi.org/10.1101/2024...

Any thoughts from people with neural network expertise? Papers describing these models often show plausibly that unsupervised training has learnt something about the data. Why is this no better than random for downstream tasks?

I also find a day of immersive reading quite refreshing. Curtin library is very peaceful at the moment.

🚨Out now 🚨😎

Our latest paper published in @CommsBio led by the
@jameskhane.bsky.social lab @theccdm.bsky.social shown the ⚖️balancing act of RIP vs negative selection in generating genetic diversity in a major🌾wheat fungal pathogen!

Find out more! 👇👇👇

A few years of practice later...

John Hawkins: 'I was a boy from out the back blocks who took on the giant' WA agricultural industry legends awarded Australia Day OAM.

So proud to say my uncle John Hawkins got an Order of Australia medal. Arriving as a child migrant, he became an important figure in ag, taking on the Australian Wheat Board to build a key market for growers.
Full story here, requires Farm Weekly sub: www.farmweekly.com.au/story/887534...

That's so skibidi 😞

But what if you were in a hurry and you wanted to tell someone you posted on a specific platform? That happens all the time for me ;)

I like that. Sounds like the start of a jazz scat.

Hi. I'm on Bluesky. What do you call Bluesky posts? Bleets? This is my first bleet.