Great work as always! Can’t wait to try it out. Whats the explanation for superior enterprise performance, at least for IDs?

Could be wrong but I just checked and pretty sure starting postdoc salary at UCPH is only up 14% since I started my position in 2017. Cost of living is up like 40%? Brutal out there

From what I know they are successful but not cheap. At least based on the quotes I have 😇

Our iTims3GS is still going strong so we’re going to hold out for the 10 year anniversary model

In all seriousness though - and to Bruker’s credit - the possibility to do field upgrades is pretty great

Just found an absolutely epic email from a shared MS facility ca. 2015 😂

True words of wisdom. #TeamMassSpec

Only time it helped me was when I broke some bones a week before a major grant deadline. Had to finish it with one functional hand. Got the grant.

I feel gross when I do it with busy work though. Which I still do too often.

Easter long weekend activities 🇩🇰

Sheesh! Thankfully I didn’t have that but I did have to go in for a crisis meeting about my next contract while on parental leave when my daughter was a week old. Was hard to forgive that one.

Such an amazing feeling when a 2-3yr contract lands in academia. Suddenly you see a big runway ahead of you. Really dislike the anxiety of the last year of a contract though. In my 10yrs post-PhD I’ve never known more than ~3mths in advance that I would have another contract.

Latest developments and tools for data analysis EuBIC Seminar 2026 Event Schedule for Wednesday the 15th April 2026 Location is in the Mærsk tower at Panum on the 15th floor in the seminar room 7.15.92 (Foredragssalen): Take the elevator to the 15 ...

On the 15th of April we will host a seminar as part of our ProteoBench hackathon in Copenhagen. Swing by and hear more about Protebench, EuBIC and other interesting projects!
Find details here: eubic-ms.org/events/lates...
See you there! 😄

1) Proteomics has a severe problem.
We keep pretending database search is “good enough”, while systematically missing everything that isn’t already known.

No genome/proteome → no ID
Variant peptide → invisible
Novel biology → filtered out

There’s also a bit of the issue of statistics for single peptides vs multi-peptide proteins. Quant and detection robustness is obviously way better at protein level. Not that it’s not possible but I’ve tried (and mostly failed) to do quite a bit of proteoform analysis in bottom up tryptic digests.

Never ceases to amaze me what you can find when you know what you are looking for (weird and wonderful PTMs, splice variants etc). For me I have to admit I have the impression DeNovo still isn’t mature enough for large scale analysis. My other guess is the complexity of data analysis/interpretation.

MR-SP2: A Microreactor-Based Workflow for Few-Cell Spatial Proteomics on the Legacy Zeiss PALM MicroBeam Laser capture microdissection (LCM) combined with liquid chromatography–tandem mass spectrometry (LC-MS/MS) enables spatial proteomics at the few-cell level but is constrained by cumulative losses during specimen capture, surface adsorption during processing, and sample transfer prior to LC-MS/MS analysis. The capture-associated losses are particularly relevant for pressure catapulting systems such as the legacy Zeiss PALM MicroBeam, which, despite discontinuation, remains in active use and therefore requires compatible low-loss workflows. We present MR-SP2 (microreactor-based sample preparation for spatial proteomics), a one-pot workflow integrating reproducible Zeiss LCM-cut specimen capture, processing with minimized adsorptive losses, and pipetting-free transfer with Evotip disposable precolumns. The workflow was evaluated using a formalin-fixed paraffin-embedded (FFPE) murine kidney tissue analyzed by timsTOF flex LC-MS/MS analysis. Across 50,000 μm3 regions (22 cells), MR-SP2 modestly improved proteome depth (3381 ± 80 versus 3174 ± 59 proteins). Decreasing sample input further accentuated the advantage of MR-SP2 in maintaining higher identification rates, highlighting the successful reduction of the adsorptive losses. At 12,500 μm3 (5–6 cells), identifications increased to 1145 ± 188 versus 302 ± 126. At 3125 μm3 (1–2 cells), identifications reached 695 ± 112 versus 206 ± 51. MR-SP2 improves identification depth for few-cell FFPE samples by nearly 3-fold compared to conventional tube-based workflows.

Nice work but please if you’re going to try LCM with the PalmRobo just capture + process directly in a PCR cap either open in a humid chamber or with them placed on top of a 96-well PCR plate. Then acidify + spin into evotips or homemade stagetips.

pubs.acs.org/doi/10.1021/...

Unsolicited advice always welcome! I have such a strong preference for vector graphics but trying to break that for the LLM adventures

Hehehe yeah Codex gets a bit loose at times. Plots are always the most questionable.

Latest agentic coding adventure:

Lightweight rust based contaminant QC tool to process directly from bruker .d files (with inspo from HowDirty and RawTools). 5s per file runtime on my Mac with SSD.

Janky but does the job.

That’s great! Actually needed that for something recently and had to live with a library.tsv so good to know I can get it at PSM level

Woo! Well done 👍 Giant files I guess?

Not sure TBH but I feel like there has to be some way to export it

I think you can do it if you export tsv pepxmls? I guess library.tsv has that info too but then it’s a consensus precursor ID rather than per sample/run

Pretty hilarious to use the fact that collagen is missing an essential amino acid as a selling point

“Look how great this is, we have 8 out of 9 essential amino acids!”

So the EU and Australia have finally reached an agreement on free trade. Let the Vegemite flow!!!

Seriously though… our supplies are at critical levels. Send help.

Yup confident it’s not ToF/detector saturation. When we up the accumulation and ramp times from 100-150-200-300ms the effect amplifies. I’m guessing the +1 ions are occupying a different ‘space’ in the trap, or are less affected by charge repulsion/heating?

I guess ICC 2.0 does something like this? For us we actually kinda really want the +1 peptides. As I understand it ICC on TimsTOF Pro (our instrument) would penalise them by lowering accumulation times because of the abundant +2/3 cloud.

On the flip side - I have a great ‘method’ for enriching +1 ions 😂

My thought exactly. I have a quote for that 🫠

TimsTOF people / people who understand ion mobility devices - is it normal that +2/+3 ions overload the trap before +1 ions?

We are observing that both +1 IDs and intensities keep going up with increasing loads but +2/+3 saturate quickly.

Had to leave early this morning to give a presentation north of Copenhagen at the SSAR conference in Krogerup. I know it’s going to go well because my son (5yo) made me a snack for the way.

Yeah I’m pretty sure it’s conformers. Just keeps me up at night, you know?