Perfinity. Remove the variability from your digestion. Oh..and do it in 4 minutes.

I think you could very successfully argue that sample prep may be the aspect of proteomics where the most chaos is introduced.  This is particularly true if you are accepting samples from multiple sources.  Digestion is one of those steps.  How much detergent (if any), urea (if any), reduction and alkylation techniques (and on and on) are all highly variable.  I'm not sure I know two people who do this exactly the same way.  And even when you do this the same way, there is still variability (see this very thorough analysis this year from Piehowski, et al., from May of this year).

Perfinity is a company in Indiana that wants to remove the variables from your protein digestion. They do this by employing a thermocycler and some very cool and very proprietary techniques to obtain reproducible and incredibly thorough digestion and peptide map coverage.  The technique, called, flash digestion can digest up to 96 samples at once and has the potential for creating these digestions in as little as 4 minutes.

For more information, see the Perfinity website.

New quality control tools for proteomics!

A few years ago I first submitted my book proposal for a text detailing Quality Control in Proteomics.  One of my favorite rejections involved a statement about a general lack of interest in the field.

But now!  QC is all over the place.  We have cool programs like SympatiQCo, and awesome reagents like the peptide retention time calibration mixture (PRTC).  More and more people all the time are using some level of quality control before just shooting their samples onto their instruments.

Now we have a new resource for targeted studies of human plasma, namely commercially available protein quality control standards.  These are being produced by a company called MRM Proteomics and you can read more about them here.

The kits from MRM proteomics are heavy labeled peptides from known biomarkers.  It is important to note that they are single peptides from these proteins of interest.  They are going to be extremely useful for testing the general health of your LC-MS system (like the PRTC peptides).

They may also be useful for setting up preliminary assays for these proteins of interest.  An important note:  the ASCP and other clinical organizations will not accept a single peptide from a protein as a positive finding for any assay.  Most reviewers will not accept a single peptide either.  But for setting up your experiment and for making sure your LC-MS is in tip-top condition, this is going to be another extremely valuable tool for our labs.

U.S. Shutdown is over! Let's do some proteomics!

What good news!  The shutdown is over.  Time to flush air, purge solvents, calibrate, and get back at it!  This has been a particular kicker for me, as a good friend of mine is on maternity leave and she left me open access to her lab while she's out.  I got one fun full day of experimentation before everything locked down.  This morning will be frantically rearranging my schedule to get back in to 1) get my things I left there because I never thought the government would shut down and resume the experiments I had started and 2) try to get in to all the people at the NIH that I had planned to work with over the last couple of weeks.

Do you need nanospray to do good proteomics? (Part 2)

I threw out this idea a while back, and some recent experiencies really make me want to reapproach it.

This week I worked with a core facility in Pittsburgh (which is quietly becoming a mass spectrometry powerhouse of a city, btw!).  At this facility, the majority of the work coming in is small molecule work, though nucleotides, whole proteins and shotgun work does make appearances.  Due to the chemical nature of the work, the HPLC in use is a high flow Ultimate 3000, and not enough shotgun proteomics is coming (yet) to justify the purchase of a nanoflow LC.

No problem.  Using the Peptide Retention Time Calibration mixture as a standard, we first benchmarked the sensitivity of the system (a Q Exactive with a 15 cm x 2.1 C-18 column, HESI source and 200 uL/min flow rate).  

I know this is hard to see, but this is 1 picomol of each of the 15 PRTC peptides.  The TIC baseline is almost 1E8.  At 1 picomol.  All 15 peptides came off nicely, when they should have.  The chromatography, obviously, could be improved, but our goal was to benchmark our sensitivity.

Next came the real sample.  A group down the hall digested a mouse liver (in solution), desalted and brought over a vial.  We set up a quick, first pass run to go overnight and came in this morning to this beautiful TIC.

Again, I apologize for the grainy JPEG.  The basepeak signal intensity is around 8E8.  These are quick runs.  Little time was spent optimizing the source conditions, dynamic exclusion, fill times, etc., we just injected a few different size aliquots of this digest.  The digest was 200 ug of mouse liver protein digested in solution and desalted.  If 0% loss, the above injection is 10 ug of protein.  Considering the losses involved due to membrane proteins (liver has an awful lot of membranes in it and no detergents were used) and to desalting, I would be surprised if we wee looking at 5ug of protein.

Yes, 10 ug on a nanocolumn is a lot.  But, come on!  10ug of protein is nothing for most biologists. You get 1-2 mg of protein from a T-75 flask of poorly growing adherent cells without even trying.

How are the results?  Using default percolator outputs provided, ~5550 peptides  and ~1250 unique protein groups.  You can't tell me that isn't awesome for a first pass, non-optimized 80 minute run.  You can, I guess, but I won't believe you.

Next, we took that run and exported a Q Exactive exclusion list through PD and put that in for an otherwise identical rerun.  Summing the two run resulted in ~1650 unique proteins and ~7000 unique peptides.  Not too shabby, in my opinion!

This experience was reinforced when I spent some time working with a big company recently.   Although I saw dozens of mass spectrometers doing proteomics experiments, I never once saw a nanoflow system.  I think that experiments of this kind are becoming more of a regularity than a novelty.  And the work this week demonstrates why.  Mass spectrometers are sensitive enough to work without the added sensitivity from nanoflow sources.  Unless your really need to be digging into the noise to look for the lowest copy number peptides and PTMS, electrospray or microspray may be enough to get you the the proteins that you need.  Nanospray will always result in higher signal, but sometimes you have to take a step back and think about just how much signal you really need.

BTW, these screenshots were graciously given to me by Dr. Bhaskar Godugu, Director of the Mass Spectrometry Facility (Chemistry) at the University of Pittsburgh.  

Spectral libraries vs. search engines

Wow.  That image is terrible.  Ugh...

Anyway, it's the content that matters!  And this topic is going to be cool, and important in the future.

Spectral libraries are something we're hearing more about.  That's because the libraries are getting bigger, more useful, and improving in quality.  I think it is a good time to take a look at what they are and what advantages/disadvantages they are bringing for us.

Spectral library searches have been around for a long time.  Originally, they were used for small molecule searching, but they were quickly adapted for peptide searching.  The two that are integrated into Proteome Discoverer 1.4; MSPepSearch (NIST; link coming when the government reopens...) and SpectraST  were two of the earlier algorithms to pop up.  

The concept is simple -- you take identified spectra that you (or somebody else) has sequenced and identified with high confidence in the past and you put that in a library.  Then on your next experiment, rather than go through the statistical magickery of a search engine, you simply compare all of your MS/MS spectra to that of your library.  If the new spectra looks like the old spectra, you have a match.  

PROS

Faster.  Way faster.  In the original paper for SpectrST (by the way, I just found out today that this rhymes with "contrast"), on the same PC, the spectral query speed for SpectraST was 0.005 seconds, while Sequest was 6.4 seconds.  That is almost 1300 times faster.  Partly this comes from the fact that you are comparing a spectra that actually occurred to another.  In a Sequest search, the engine has to look at every possible MS/MS fragment ion and do that comparison.

More sensitive.  By comparing two spectra, you can get away with fewer fragments and of lower intensity than you can with a traditional search engine, mostly for the reasons mentioned above.

CON

You've got to have a library.  And an okay library only cuts it if you want okay results.  If you want good results, you need a good library, and excellent results...  If your spectra has a PTM, but that PTM has never been recorded in a library, that result is gone.

And this is why I haven't really used these engines.  The libraries just haven't been good enough.  But this is the good news:  They are getting better.  Much better.  High resolution MS/MS libraries are around the corner and new tools are coming.

But this is the best news:

With the completion of all these new libraries, we won't have to worry about whether spectral library or traditional searching is better, because we can use them both.  We can use the spectral library engine to filter rapidly though matches, then we can take the spectra that don't match and send those (and only those) through Sequest, Mascot or another engine.  Then, if you really want to you can take the spectra that don't match there and export those by searching with Byonic or Peaks or PepNovo+ and really get down into your data.  

For a video on how to set up this last part, follow this link (watch in HD only!)

Tandem mass spectral libraries for phosphopeptides

Analysis of the proteomics data using spectral libraries has become possible but with the fact the unavailability of a comprehensive spectral library. Although, it is evident that such libraries will become more popular in coming years.

This technical note by Henry Lam Lab just accepted in JPR describes about the spectral library generation for phosphopeptides from human as well as four other model organisms. The claim is better sensitivity over conventional database searching to identify phosphorylated peptides. The other good part is that this library is made freely available to be searched using SpectraST algorithm (PD 1.4 has this algorithm in-build for spectral library searching).

I will be testing this phospho-spectral library and see how it performs in my hand.

Proteome Discoverer International User's Meeting 2013!

Thermo Proteome Discoverer Users’ Meeting Agenda

Courtyard Marriott, 777 Memorial Drive, Cambridge

Wednesday, Oct 23, 2013

08:00 – 09:00 Registration and Breakfast

09:00 – 09:15 Welcome

9:15 – 10:00 PD 1.4: Overview of New Features; Bernard Delanghe, Thermo Fisher Scientific


10:00 – 10:45 Byonic and Preview: New Tools for Proteomics and Glycoproteomics Analysis;

Marshall Bern, Protein Metrics Inc.


10:45 – 11:00 Break


11:00 – 11:45 Using MS Amanda for Identifying High Resolution and High Accuracy Tandem Mass Spectra;  Viktoria Dorfer, University of Applied Sciences Upper Austria, Campus Hagenberg


11:45 – 12:30 A Graphical Bayesian Approach to Mass Spectrometry-based Protein Identification
Spectra;  Oliver Serang, Thermo Fisher Scientific


12:30 – 13:15 An Automated Tool For Creating and Using Curated Spectral Libraries; Barbara Frewen, Thermo Fisher Scientific


13:15 – 14:15 Lunch

14:15 – 15:00 PD 2.0 preview;  Bernard Delanghe, Thermo Fischer Scientific

15:00– 15:30 User presentation TBC

15:30 – 15:45 Break

15:45 – 17:00 Q&A session and wrap-up

17:15 – 18:00 BRIMS tour

Thanks to some last minute scheduling changes, partially brought about by the Government shutdown, I will get to be there.  To register, follow this link.

Thorough analysis of the most influential authors in proteomics

I found this really nice blog through a twitter link the other day.  I strongly recommend that you check it out.  A cool entry is a breakdown of the number of proteomics/genomics/bioinformatics publications over time, as well as the number of citations the big guys are getting.  What is interesting/unfortunate is the complete lack of women on this list.

Iterative searching in Proteome Discoverer!

Want to boost your peptide IDs in Proteome Discoverer as well as the confidence of those IDs?  Move from your regular search to iterative searching.  I looked it up (it·er·a·tive ( t -r t v, - r- -t v). adj. 1. Characterized by or involving repetition, recurrence, reiteration, or repetitiousness ).

Here is the gist of it.  You take that file and you give it different instances of search engines with different combinations of likely modifications.  For example, if your peptide is one that has acetylation and phosphorylation, you need to have both of those mods in that search, but it is impossible to have every combination (or even very many) in one search engine.  Using multiple engines allows you to search more modifications.  It also allows you to get better confidence on your ID'ed peptides.  For example, if Mascot, Sequest and MSAmanda all give you a peptide spectral match, then it is probably more likely to be true than if only one of the engines give it to you.  So your confidence increases.

How are the results?

On each sample, this iterative (comprehensive search) increased the number of ID'ed peptides.  By a lot.  Yes, at first it seems crazy, but it sure looks like it works.  For full information, look for this on Planet Orbitrap:

DanteR: Data Analysis software by PNNL

Pacific Northwest National Laboratory (PNNL) has developed many useful and handy software tools especially for proteomic data analysis. DanteR is one among them which uses R programming environment. It is user-friendly with graphic-user-interface and helpful to do many of the common tasks such as normalization, Annova, clustering to name a few. It is easy to install and intuitive. The best part is if you are a programmer, you can add your programs to it using the add-on feature. You can download DanteR from here.