Kevin Welle

Yes sample prep for MagNet is very important. We noticed a huge decrease in IDs using serum instead of plasma. Anecdotally, even relatively small changes in centrifugation speed when the blood is collected seems to make a difference.

The Mag-Net data surprises me. We’ve used it for several projects now and usually get 3-4k proteins. We are also using an Astral but with only a 12 min gradient (20 min injection to injection).

Orbi res doesn’t matter at all (or barely at all- there is slightly more overhead between an Orbi and Astral scan that when doing just Astral scans). We do a 240k res MS1 scan which takes 0.6 seconds, so that’s what we set the cycle time to.

Website hasn’t been updated in a while. 72 SPD run is $140 for internal customers. But thanks for the reminder to nudge IT again to get the new instrument and rates on the website.

No we pack our own columns and use a trap as well. Since getting the Astral we pack 15 cm 100 um ID columns with 1.8 um beads and flow at 600 nL/min. Accuracy was quite incredible when testing a 2 proteome mix at different spike-in levels. Probably getting about 5-6 DPPP from base to base.

@WelleKevin on twitter. I unfollowed almost everyone on there and just use Lists now which makes it much better for me. My DMs are open there though.

Happy to talk in more detail via DM on twitter if that would be helpful.

My guess is 6k proteins is pretty trivial at 100 SPD. We usually run at 72 SPD (20 minutes injection to injection- about 10 minutes of which is active gradient) and get about 8500 proteins from HeLa. We actually just got 10k proteins from a 72 SPD method with astroglia cells.

We routinely use DIA to analyze pull-downs. I try to avoid TMT most of the time, but regardless I think our DIA results for pulldowns have been great.

I just spoke to a collaborator in Denmark and he said they use World Courier for overseas shipping. Apparently they re-supply dry ice during shipment and packages arrive in 2-4 days. Maybe worth looking into.

Proteome birthdating reveals age-selectivity of protein ubiquitination bioRxiv - the preprint server for biology, operated by Cold Spring Harbor Laboratory, a research and educational institution

Thrilled to share my first pre-print !! www.biorxiv.org/content/10.1...

Could use the Astral for this. Keep doing continuous 500k resolution scans in the Orbi while the Astral is handling all MS2 scans. That’s pretty much how the Astral runs now, but the default MS1 res is 240k.

We’ll hook you up with one. We are doing a bunch of testing this week. If we do some runs on low-input lysates with multi-CV FAIMS we can send you a raw file to play around with.

Won’t really be able to use FAIMS for normal DIA runs on the Astral as switching between voltages takes about 25 ms. We’ll be using it primarily for TMT and for single-cell/low-input samples where injection times will be higher.

And maybe TMT experts can correct me if I’m wrong, but I don’t think TMT will be able to drastically shorten gradients or get away with not fractionating. Isolation interference will always be an issue if compressing those gradients or increasing sample complexity by not fractionating.

We actually would do 2 hour runs on our 8 TMT fractions because we found SPS-MS3 was so slow, and to reduce co-isolation interference. With DIA runs becoming so short with next-gen instruments my instinct is that even with 16 samples DIA will use far less instrument time than a fractionated TMTpro16

For these low abundance peptides, don’t co-eluting ions have a disproportionate impact on getting accurate quan values? When doing a 3 proteome mix with known ratios, it was striking to me how poor the accuracy was for low abundant proteins. This was on a Lumos with SPS-MS3, but without RTS.

Maybe we just didn’t spend enough time on optimizing TMT settings on our Lumos, but we got much better data with DIA than TMT. This was with a 3 proteome mix with known ratios. DIA identified far more proteins, and with far better quantitation. TMT samples were fractionated into 8 fractions too.

Wow that’s great, especially if a one time charge. We get charged $200/TB/year for storage that is guaranteed to be backed up.

How much does your University charge for space? We’d go bankrupt if we bought 2 Petabytes!

Make sure the freight elevators can handle the length of the crates. 7 feet long and we needed to go through multiple different elevators and hallways because of where our lab is. Also the crate for accessories is almost as big as the astral crate.

We actually had 3 guys delivering it. Thank god because it was perilous to get it through the hallways and the multiple elevators. Two massive crates. Took almost 3 hours total to get both outside our lab.

I didn’t want people to think I was posting about a new floor centrifuge! Install should happen on the 16th if all goes well.

Look what just arrived! Can’t wait for this to be installed in a couple of weeks! #astral #teammassspec

How long does it take to re-equilibrate those columns? That is 4.4 uL of column volume, and using 5 volumes for equilibration would bring it up to 22 uL. Are you re-equilibrating at 1000 bar so it only takes a couple of minutes? I assume the normal flow rate for that column is around 1.5 uL/min?

You are not alone. It’s like wtf went wrong with the BCA/digestion/labeling where there could be 2x more protein in one sample vs another?

@neely.bsky.social sent me one already! Thanks for the repost.