Moving on

Today is bitter sweet.  One of the best scientists I've ever worked with is leaving my team today to pursue bigger and better things.  While her skills will be greatly missed and extremely difficult to replace, I'm excited by the fact that we'll probably get to hang out more often since people generally don't need two Field Apps Scientists in one place and I tend to spend quite a bit of time in Boston.  And probably more now, since she won't be there to take most of the requests in the area.

Good luck with this next stage, Dr. Cardasis!!!!  And don't think you're going to get out of giving me lessons on instrument physics any time soon!

SparseQuant

Classical thinking says to do peptide quan right with a mass spec, you need a labeled peptide for each and every peptide that you are doing quan on.  The problem is that this can be crazy expensive.

Label free quan doesn't use require any fancy expensive labels, but the quan is relative and can be difficult to reproduce.

What we need is a compromise.  What if you combined the technology to make it expensive AND difficult to reproduce?  Just kidding!  Would that be a paper good enough for Nature Methods?  Probably, but this is even better.

This new paper in Nature Methods demonstrates the power of integrating just a few labeled peptides and using them for obtaining an improvement in your quan.  Cut the price, but still get better quan than you will see from label free. All the benefits of targeted quan with a whole lot less price!
             

Proteomics in personalized medicine

I just learned about these companies this week and I'm pretty excited about them.  These are both companies that are using proteomic biomarker assays to supplement personalized medicine.

Integrated diagnostics is based in Seattle and has a system called XPresys Lung that uses biomarkers to determine whether lung nodules are cancerous or benign.  Check it out!

OncoPlex DX is based in Rockville, MD and has tests for biomarkers for lung, breast, and colorectal cancers. Their website is here.

 I LOVE seeing proteomics moving toward the clinic.  So many good people are out there using global screens to find biomarkers and here are people using those biomarkers to benefit patients?

Proteomics WIN!!!

CompOmics

I ended up at a bar restaurant with some awesome scientists last night and one of the many things we ended up talking about was the number and quality of free resources that are popping up in proteomics.  Its getting to the point now where just about everything you'd want to do is out there.  The trick, a lot of the time, is in hunting down the software that will do what you want.  Unfortunately, the other trick is in getting it installed because sometimes our bioinformatics friends forget that not all of us know R and Perl, or use Linux.

Fortunately, this later problem seems to be getting better and I especially expect it to improve now that we have a precedent established of publications based on changing great freely available programs into easy to operature (and install) GUI programs.

To the topic at hand:  More free resources!  I was alerted to this one by a reader in India (thanks Dr. Sreelakshmi!!)

The CompOmics group in Ghent, Belgium has a great website with all sorts of free software as well as tutorials on how to use said software.  The software ranges from data parsers (from OMSSA or Mascot) to cell motility analyzers.  In short, tons of cool stuff!

Direct link to the site here!

Plos One new data access policy

As of March 1st, all Plos One journals require a new section at the end of the article you submit.  This is the data availability statement.  All authors must not only make all data available, but it must be specifically stated where this data is and how to obtain it.  For more information on the new Data Policy, click here.

ABRF 2014!

Woooooooo!!!!!!

N15 labeling in Proteome Discoverer

This question pops up a lot, despite the fact that this workflow is available on both the BRIMS portal and PlanetOrbitrap.  It never hurts to put it up in an additional place, right?

Question:  Can I do N15 labeling quan in Proteome Discoverer
Answer:  Yup!  But its a little tricky.

This link will take you to a Powerpoint with some instructions and an XML file that has a method I made for someone for PD 1.3.

Ultimately, the real trick is adding the modifications you need for each amino acid.  You need specify these mods uniquely by how many nitrogens are in each of your amino acids.

First start by updating your Unimod database.

Then add these labels.

Once these are in there, you can basically build the method based any of the precursor based quan methods (like SILAC). Comparing your light to your heavy.

Now, it gets a whole lot trickier if you are looking for incomplete incorporation.  You begin to run outside the realm of what you can do with the SILAC node.  What you need to do at that point is run your N15 mods as dynamic and use the precursor ion area detector and some database work will be necessary.  It certainly isn't impossible, but it will make for a nice long day of analysis.

The XML method in this file assumes that you are trying to compare two things, channel 1 with no N15 and channel two with 100% incorporation of N15 into each amino acid.  If you alkylate your cysteines you will need to make a custom modification that includes the mass shift of your N15 on cysteine as well as the mass shift of your alkylation (cause we can't yet do 2 mods/amino acid).

EDIT: Here is the new download link for the PD 1.3 slide deck.

More evidence -- DMSO study on the Q Exactive!

Early this morning I received an awesome Powerpoint describing results of another DMSO comparison.  This time it was for a TMT experiment performed on a Q Exactive instrument.

The work is by Dr. Henrik Johansson and Dr. Rui Branca of the Karolinska Institutet and my colleague Dr. Michaela Scigelova.

Guess what?  DMSO additives in buffer reproducibly increase peptide/proteins in the QE during nanospray applications.  Interestingly, it isn't to the same degree as the increases we have seen in the hybrid platforms.  I obtained permission to distribute the document this afternoon, so I'll put a direct link here until it is published on Planet Orbitrap.

Highlights:  
Adding DMSO led to 69% increase in +2 charge state peptides.  This is particularly cool, because TMT labeling tends to increase the peptide charge state (which can translate to a decrease in sequencing effiicency)

Adding DMSO led to a 17% increase in unique protein IDs.

Interestingly, comparison of the same sample ran twice, DMSO vs. normal buffer = new peptides!

Makes me wonder what those of you with super fancy multiple pump nanoLCs could do with this information....

Now, as always with this buffer additive -- I want to strongly caution you.  We do not have longitudinal data yet to fully realize the implications of this (or any other) buffer additive on nanoLC pumps, seals, or even the MS system itself.  Please follow the guidelines and specifications for your particular LC, source and instrument.  When in doubt, always please contact your respective vendors and technical support lines.  Also, please refer to my Disclaimer statement regarding this and any other advice you obtain through this blog.

You can download this incredibly informative and thorough analysis here.

Science fair...

Had to share this.  I'm guessing the parents aren't scientists....

Antibodypedia!

I was reading through JPR abstracts and saw the mention of this resource.  The article, however, wasn't open access, so they won't get any credit here.  (I don't want y'all thinking I read something that I really didn't, anyway.)

This resource is awesome and I wish it was around when I had to "validate" every proteomics result with some sort of antibody-based assay.

Here is how it works:

1) You enter your protein of interest and hit the search button

The antibodypedia comes back with some ontology information on your protein, a list of orthologs and how many antibodies are commercially available.

By clicking on the antibodies that are available, you get an expanded report leading you right to said antibodies!

Not only do you get the provider AND the catalog number, but you get an easy reference table for what that antibody is useful for; Western, immunocytochemistry, precipitation, and on and on!

The solid green dot means that the antibodypedia has verified evidence that the antibody works for that type of assay.  The 3/4 dots means the manufacturer shows evidence that the antibody works for that assay, but it hasn't been individually confirmed.  The 1/4 dot means that the manufacturer says it will work, but there is no evidence out there.

How easy is that?!?!  I've been a little uninspired this week so far, but this is exactly the kind of stuff I want to write about.

You can directly link to the antibodypedia here!

Free seminar on pesticides via LC-MS

Free WebCast on Tuesday!  You can register here!

Details of the Thermo -- Life acquisistion

GEN has this great article giving a nice overview of the Thermo acquisition of Life Technologies including what it is keeping and what it is selling to GE as part of the deal.

Worth a quick read!

Thanks @servingscience for the link!

Want to boost MRM sensitivity on a ton of peptides at once?

I love the cross-talk happening in the field right now.  Our counterparts in metabolomics and in targeted quan have tons of cool chromatography columns and derivatization techniques that make our C-18 reversed phase stuff seem kind of lazy and unsophisticated, though some cool stuff (like HILIC) have crossed over and are showing their uses.

Well, what if we have stuff to contribute back to the targeted triple quad game?  It seems like we do.  A new paper in press (and currently open access, quick grab it!) from Steven Carr's lab at the BROAD (like Toad) shows the usefulness of one of our chromatography tricks in targeted peptide quan.

What's that trick?  Long columns with small particle size.  Heck, its the only LC trick we have, right?  The paper, aptly named "Simplified and Efficient Quantification of Low Abundance Proteins at Very High Multiplex by Targeted Mass Spectrometry" demonstrates the value of this trick on increasing both the number of peptides quantifiable in a single run as well as an improvement in dynamic range.

Oh yeah, and its a holiday.  Here is a great poem for it.

Who is doing what drugs and when?

Again, not proteomics, but interesting!  People all over the world take drugs.  In most of the developed world these same people use bathrooms connected to centralized municipal sewage systems.  Well, if you didn't have some weird aversion to sewage and you had a super sensitive mass spectrometer, you could totally find out who is doing what drug and when.

Researchers all over the world have done just that.  And the results are really interesting.  This recent article at "The Verge" has a summary of the results of some of these studies.

Many labs use triple quad MRM for analysis of samples as dirty as raw sewage.  Biljsma et al., describes a study from 2012 where these Dutch researchers showed that they can do the exact same (or better) analysis with an Orbitrap.

Interesting findings?  The Dutch don't appear to do any more drugs (except cannabis) than any other country.  Having the Super Bowl in your city does not increase drug consumption, and at universities all over the world, amphetamine levels spike in sewage during finals week.

EUPA International Proteomics Tutorial Program

EUPA has started a program directed at providing free resources for scientists (or wanna-be scientists students) to learn proteomics.

This is great news unless you are someone who likes getting a little check each month for an intro to mass spectrometry book you wrote several years ago.  Unlike this hypothetical old book, however, the free information from EUPA is current, available all over the web, written by experts, clear and concise!

For more information on this, such as how to get these nice tutorials (or possibly even how to contribute content!) follow this link.

How does varying ETD times affect fragmentation of intact proteins?

I swear, I'm not changing this into a blog for intacts and top down.  This just happens to be the only thing I've been doing lately.

This is really cool, and brazenly stolen from some ProsightPC training material.

Same protein, different amount of time that they get to react with the ETD anion.  The end MS/MS spectra are completely different!  I guess the 2 ms and 4 ms spectra have a lot of "incomplete" fragmentation points?  But looking at it from that perspective it sounds like 2 ms not only ends dropping your cycle time but gives you more and better coverage.  The 50 ms is what I'm more used to seeing, both in method setup and in sequence coverage, nice fragments from the C- and N- termini.

Worth keeping in mind, at the very least.  In terms of ETD, more may not necessarily be better (or multiple MS/MS events with different ETD reaction times may be a great way to get a shit ton of data!)

Shoutout to David Horn for providing me with this (and a slew of other) slide(s).

Intacts and top downs and bears!

What did we do without the Internet and all these resources?

Currently, intact and top down proteomics is a big focus for me.  I'm trying to dig up some primary research to get a better handle on it and everything and there are resources all over the place!

The winner (so far) for aesthetics?  TopDownProteomics.org

An extremely nice site with all sorts of tutorials, video resources, software descriptions, new papers, etc., etc.,

If you haven't seen it, head directly over to it here.

Sorry, I lied about the bear.  And I meant to post date this one to show up later in the week....oh well....

While I'm editing, I might as well add a bear...

QE Plus intact antibodies

I get distracted sometimes.  Months ago, I promised to show some more data on what the QE Plus can do with intact proteins, and here it is, finally!

This is the 0.1 ug of the Waters antibody standard injected onto a C4 column running at 300uL/min.  The S/N is >100:1.

Post-deconvolution, we are looking at sub-10ppm mass accuracy.  I ran this, btw, not some intact expert!

Punch line?  This is without gas optimization.  This is also without enabling "protein mode" on the QE Plus. This was without desalting the notoriously gross Waters mAB standard!  This was a test to make sure the instrument was working correctly.  Could I have adjusted gas pressures, optimized (thrown in a nice high lock mass?) and employed Protein mode to improve this data?  Of course!  Is 1000:1 S/N possible on a QE Plus on an intact (and not desalted, LOL!) antibody?  I don't know, but if this is what one can do right out of the box, imagining what someone who knows what they're doing and takes the time to really optimize everything is a pretty exciting prospect!

In case you are interested, the method file is now up on the Orbitrap methods database.

RNAseq plus Proteomics. It's coming!

RNAseq "next gen sequencing" data is becoming more and more common every day.  The instruments are getting better, cheaper and faster all the time.  From what I'm seeing and hearing, I'm expecting to see a ton of posters and papers out there where shotgun LC-MS/MS data is searched against RNAseq data.

The big problem?  The size of the two datasets.  Especially the RNAseq data.  It's pretty tough to search against a 500GB genomics file with our LC-MS/MS data.

Sunghee Woo et. al.,  to the rescue!  In the paper "Proteogenomic Database Construction Driven from Large Scale RNAseq Data", this team from UCSD (with some help from Mike MacCoss) demonstrates how to efficiently use the two technologies synchronously.

They start by dramatically reducing the RNAseq data from a 400GB output down to a 400MB FASTA file, by a stepwise removal of redundancy and less useful data.  The logic is clearly defined and seems really smart.

Next they compare the two datasets, the LC-MS/MS data and the new FASTA in such a way that they end up complementing each other.  So, not only do you they get better matches for their proteomics data than they would with a fully annotated FASTA (what most of us want to use this for), but they also improve the annotation of their genomics data.  That's right.  They use the proteomics to correct the genomics data.  Cause the proteomics data can give you single amino acid substitution and splice variants and improve the understanding of the genome.

Get your LC method out of your Thermo raw data files

Shoutout to my friend Dave who suggested I write this.  Then I forgot.  Then I remembered again!  Have you ever looked through your RAW data and really wanted to know more about your LC conditions (by more, I really mean anything?)
 Open your RAW data file in Xcalibur.
 Right click on your specra (after thumbtacking it)
 Open your Instrument method
 Hit the right arrow.

BOOM!!!!  Your LC method appears in the box!

JPR special issue on the C-CHP

I love when stuff is open source!  And this is hands-down, one of my very favorite projects out there.  You guys working on the HPP get all sorts of love from me.  The C-CHP, especially, since the common theme seemes to be high resolution accurate mass spectrometry and every study so far seems like it could be subtitled "how to get extremely deep proteome coverage with an Orbitrap".

Want to see how this is going right now?  Want to see how effectively scientists can work together all over the world to answer basic science questions?  Check out this special issue at JPR!

Protein Deconvolution tutorial videos

I like this banner.  I don't know where its from, and I'm probably not allowed to use it, but Google Images gave me it instead of what I was looking for.

Anywho!  One of my spare time projects the last couple weeks has been making new tutorial videos for you guys.

These aren't officially branded things.  These are rough drafts.  But maybe one day they will be a real thing.  Anyway -- check out the sidebar.  There is a new category.  Protein Deconvolution tutorial videos.  There are only 2 right now, but there will (probably) be more later.

So far there are two videos:

How to manually deconvolute a single intact antibody LC-MS file:  https://vimeo.com/85852770

How to batch deconvolute a series of intact antibody LC-MS runs:  https://vimeo.com/85855510

I have no idea why the font just changed there.  Anyway, the links probably don't work yet, but there is also a link to download the monoclonal antibody LC-MS RAW files I used for these analyses.

Reddit proteomics!

Almost a week since my last post!  I have been busy...  I promise what I've been working on is really cool and, with any luck, should appear on the side bar in the next couple of weeks.

Time to get back in it.  What happens when you combine 2 of my very favorite things in this world?  Not Pugs and Porsches.  (Though...you can't get much cooler than a pug in a Porsche...)

What about Reddit and Proteomics?  If you aren't familiar with Reddit: "The front page of the Internet" you probably shouldn't check it out.  But if you are already addicted to the fastest way to get news and awesome content about whatever it is in the world that interests you, then you should add www.reddit.com/r/proteomics

Right now there isn't much on there, but that's the joy of Reddit, the more people that get on and participate the better it gets.  Plus, if you are on #612 in your personal newsfeed, it wouldn't hurt if a couple of things in there were cool Proteomics links, right?

Oh, and if you have no idea what I'm talking about at all, check out this awesome video entitled "what is reddit?"