Vanda G-Lovejoy

[Vanda G-Lovejoy @non_coding_gal]

Finally out in NAR ! 🥳Meet ECSFinder, and ML tool that identifies the most informative conservation/covariation features for RNA structure and benchmarks well on hTERC. Grateful to @martinalexsmith @sauvagem1 & John Mattick for their contributions.

[Martin Smith @martinalexsmith]

RNA is far more than a messenger.

Its structures regulate, evolve, and catalyze — carrying out functions DNA and proteins cannot alone.

But predicting #RNA structure from sequence is still extremely difficult.

In our latest NAR paper, we present a way forward.
🧵 ⤵️

[Vanda G-Lovejoy @non_coding_gal]

@chidiacrony @stemcellreports @angerslab Bravo Rony!!!!! 👏

[Vanda G-Lovejoy @non_coding_gal]

@martinalexsmith @ppgardne The RAFSp option is essentially the RNAalifold stacking energy, albeit without the background correction, which is one of the features investigated in our study and dissected in our analysis.

[Martin Smith @martinalexsmith]

This eventuated with ECSfinder, a pipeline that can be used to detect evolutionarily conserved 2D structures (ECS) with greater accuracy than individual tools. All work led by PhD candidate @non_coding_gal with significant contributions from @sauvagem1 and John Mattick (10/10)

[Martin Smith @martinalexsmith]

The great thing about interpretable ML is that you can look at feature importance in the resulting classifier, which exposed some interesting findings. Notably, minimum free energy is somewhat important! (9/)

[Martin Smith @martinalexsmith]

Do long non-coding RNAs form conserved, functional secondary structures?🧬䷰𐂷 How can we systematically assess this?
We compared 2 tools fit for this task (SISSIz & R-scape) across 2 genomic benchmarking strategies.
Spoiler: machine learning wins (🧵/10)
https://t.co/fCxZEoI2r1

[Vanda G-Lovejoy @non_coding_gal]

@JF_CoteLab @ircm_mtl @SocMuscleBio Bravo Sarah!!!!

[The Figen @TheFigen_]

First day at work.

[Vanda G-Lovejoy @non_coding_gal]

Incredible talk on generative ai and how it can transform biomedical research by @jameszou #ISMB2024

[Vanda G-Lovejoy @non_coding_gal]

Looking forward to the next few days🤓

[Dr. Glaucomflecken @DGlaucomflecken]

False eyelashes are incredibly popular but can I just say as an ophthalmologist, they look horrifying under 10x magnification. The lashes are so densely packed that they trap all the dead skin cells, dirt, and oily secretions from your eyelids. It’s a sebaceous swamp in there and frequently results in red, itchy, swollen, irritated eyelids.

[Terese @tresterese]

I recorded nearly every* breastfeeding session with my older daughter over 1.5 years. Mostly so I could someday make some charts. Yay charts!

*I only missed a very small handful of feeds!

[Vanda G-Lovejoy @non_coding_gal]

@sauvagem1 @FRQS1 Oh wow félicitations !!

[Andrej Spiridonov @AndrejSpiridon4]

I'm glad to share our most recent article with @LovejoyShaun in the @PaleoSoc journal "Paleobiology" presenting so called Fractional MacroEvolution Model (FMEM), which reflects key features of macroevolution - scaling and long memory
https://t.co/4zmCUNbxDn
1/n

[Dr Meming @Dr_Meming]

Me after writing one paragraph of my paper (which I’ll likely delete later)

[Oded Rechavi @OdedRechavi]

Trimming the abstract to 150 words
https://t.co/Vf4a8Ujpk4

[MIT CSAIL @MIT_CSAIL]

The 12 types of ML papers.

(created by @natashajaques @maxhkw)

#MachineLearning #ML #DataScience

[Vanda G-Lovejoy @non_coding_gal]

Last day in Lorne today 🥲What a wonderful conference @GenomeConf was. Great weather, fascinating talks and wonderful people 🥰.

[Vanda G-Lovejoy @non_coding_gal]

You know where to find me between sessions @GenomeConf

[Martin Smith @martinalexsmith]

Second poster (#149) from @therealsmithlab tonight at #LorneGenome @GenomeConf by @BiologistQuirky: A comparative benchmarking of conserved RNA secondary structure prediction tools for the systematic annotation of non-coding RNAs. Ask Vanda about R-scape, SISSIz & lncRNAs