Theodoros Roumeliotis

Staff Scientist, Functional Proteomics Group, The Institute of Cancer Research, London, UK

[Theodoros Roumeliotis @Throumel]

@MariosGeorgakis Why the correlation between MS methods is so low?

[Theodoros Roumeliotis @Throumel]

Delighted to share our PhoXplex method Combining PhoX Cross-Linking with Isobaric Labeling for Quantitative Proteome-Wide Mapping of Protein Interfaces published in Journal of Proteome Research https://t.co/jqLRgqowk5

[Paul Huang Lab @HuangLabICR]

Our new paper "The Proteomic Landscape of Soft Tissue Sarcomas" is out in @NatureComms!
https://t.co/Hn1NODQ6jG

We characterise the proteomic profiles of >300 #sarcomas across 11 histological subtypes in paediatric & adult patients.

What did we learn from this study? A 🧵1/n

[Theodoros Roumeliotis @Throumel]

@UCDProteomics Combat https://t.co/Cu84IjfAvE will probably do, but if your 2 batches coincide with 2 biological groups, then the biological differences will be eliminated too.

[Theodoros Roumeliotis @Throumel]

@pwilmarth Extremely thorough and insightful! I feel that the exp design is a bit unfair for TMT though. MS time matched to what is most common for DIA (not to ~20 frcs, more common for TMT if depth matters more than speed) and use of TMT10plex instead of TMT16/18plex.

[Theodoros Roumeliotis @Throumel]

@ItsBiniR Maybe scale each protein to row mean for each cell line group separately, this would eliminate differences between the two groups and make relative differences within each group comparable between the two groups (if that's the question).

[Claudio Alfieri @calfieri84]

Very proud of sharing these beautiful cryo-EM structures of the human SIN3B complex. HDAC complexes are more specific than we ever thought! https://t.co/7xGkzwHt4t. .

[Devin Schweppe @dschweppe1]

An incredible amount of work to profile 875 compounds from Dylan & the @GygiLab! Very excited to see this out now.

[Theodoros Roumeliotis @Throumel]

@YishaiLevin I would prefer the Ionic pillar in mass spectrometry :)

[Theodoros Roumeliotis @Throumel]

An awesome poem about mass spectrometry written by AI!

[Theodoros Roumeliotis @Throumel]

@GarciaLabMS DDA + DIA = Bff

[Laurence Pearl 🔥💚🚣‍♀️🛶♊️ @laurencepearl]

Our paper is out showing how HSP90-CDC37-PP5 ‘factory resets’ client protein kinases. Great work by @oberoi_jasmeen Xavi Aran Guiu, and Emily Outwin @GDSC_Sussex with Pascale Schellenberger @SussexLifeSci and @jc4_jyoti and Theo Roumeliotis @ICR_London https://t.co/7KBoT1Lr8m

[Theodoros Roumeliotis @Throumel]

@facundofGT To give an accurate diagnosis with no shoulder peaks :)

[Claudio Alfieri @calfieri84]

Snapshot of a MuvB complex engaged with a nucleosome, first big effort from my lab is now out in:
https://t.co/wFD4KP5gF8

[Theodoros Roumeliotis @Throumel]

@pwilmarth Great read! Would it make sense to use the mean from several lots rather than changing every time?

[Theodoros Roumeliotis @Throumel]

@JohnRYatesIII @ProteomicsNews I would have gone for sub-hour, sub-quarter, sub-min, sub-sec, sub-zero scales 😀

[Theodoros Roumeliotis @Throumel]

@AJ_Brenes @pwilmarth @Peptidome @ItsBiniR I thought the same, but when I used identical digests (x3) of a control (not pool of all) across sets and scaled to their mean, still got batch effect. Maybe it is the identical nature of a pooled-all ref than any ref. But having pooled-all ref is tricky for on going collection.

[Theodoros Roumeliotis @Throumel]

@Peptidome @pwilmarth @ItsBiniR Maybe redundant info, but it is important to use two different ref samples rather than two repeats of the same ref.

[Theodoros Roumeliotis @Throumel]

Delighted to share our SimPLIT workflow for hassle-free and very reproducible large scale sample preparation for TMT proteomics! @JProteomeRes https://t.co/mTUgNcJ821