Maxquant summer school talk 2013 lectures now available

The MaxQuant summer school lectures have been loaded to youtube.  I can't seem to embed them as here as the blogger/youtube interface hasn't been updated yet, but you can find them here:

Talk by Matthias Mann.
Perseus talks by Jurgen Cox.
Perseus tutorial by Tami Geiger
Andromeda talk by Jurgen Cox
Andromeda configuration by Fabian Coscia

It appears that more videos are being posted right now.  Go to any one of these and I'm sure the clever google algorithms will lead you to the the rest.

Our field got some love from ToothpasteForDinner!

Want to learn about proteomics? Take a free course from MIT

Descriptions of the MIT OpenCourseWare project have been circulating the interwebs a lot recently.  Hundreds of courses are available -- the syllabi, lecture notes, and even the exams.  Some classes include films of the instructors actually teaching each individual lecture.

Course 6.092 is the undergraduate course, Bioinformatics and Proteomics.  The course might seem a little dated as it appears to have been prepared in 2005, but if you're looking for an intro to the field I bet it would be a good starting point.  You can find more about the course and the MIT OCW here.

Translational Proteomics: Open call for papers

In June, Elsevier's new Journal, the excitingly named "Translational Proteomics" officially launched.  The first issue is currently in production and there is an open call for papers.

I'm pretty excited by this journal, as it is open source and deals with the direct applications of proteomics technologies.  Lets stop being all theoretical and cure some diseases already!

Cell cycle regulation of microtubule interactomes: multi-layered regulation is critical for the interphase/mitosis transition

Currently in press at MCP:  the paper in the title of this post from Heather M. Syred et al., at the University of Edinburgh.

This paper is interesting for a number of reasons, not limited to the following short list:

1) Microtubules are WAY more important than we tend to think that they are
2) They are far more regulated than we tend to think they are
3) The first 2 points illustrate how little we actually know about how cells work, because these things are hugely expressed in most cell types
4) A reasonably simple method is defined in this paper for obtaining high-purity microtubules (and cytoskeletal proteins in general!) that really opens this kind of method to researchers who are very interested in cells where these structural proteins play a key roll (epithelial-mesenchymal transition, anyone?)
5) Whoever did the fluorescent microscopy did a top notch job.  My favorite way to end a paper is with IHC/ICC and this work is just stunning (and better than anything I've done!)

If you're interested in MAP kinases, cytoskeletal rearrangements, mitosis or in cell biology in general, I recommend you check this one out.

XCMS Online: Free metabolomics analysis from Scripps

The field of metabolomics has been growing at just an incredible rate.  "Explosion" is a term that I've heard more than once or twice.  I've been kind of on the outside.  Labs I've been in have prepared and sent off samples for metabolomic analysis throughout my career.  5 or 6 years ago, it seemed like you got back chromatograms with 100-200 peaks and about 25-50% of them identified within a reasonable error rate.  Now I see chromatograms that are as busy as proteomics LC separations with hundreds of identified compounds identified.

To deal with the many platforms, data files and libraries, a number of options have popped up.  This month's issue of the Scientist introduced me to XCMS Online, a free, multi-vendor supported cloud-based analysis program that seems heads and shoulders above what I've seen so far.

The output is visually stunning, seemingly easy to manipulate, and there are sample datasets to manipulate in order to learn the software.

Again, I'm an outsider looking in, but I tell you what -- I'm dreaming of the day that I can do proteomics and metabolomics in one workflow -- the kind of promise that we are seeing with the Compound Discoverer announcement at ASMS.  Until then, however, Scripps has provided us with some really nice tools to work with in the meantime.

You can read more about XCMS Online at the Scientist, or check it out for yourself here.

Transfect individual cells on a chip!

This advance is getting a lot of exposure.  It was featured in this month's "The Scientist" and "Lab Manager" and is getting circulated through the interwebs...because it is awesome.

These tiny microchannels force the cell memberane to elongate, disrupting its ability to maintain homeostasis just long enough that you can get your plasmid or whatever into them.  Talk about easy mutations!  Probably won't work on anything with a cell wall, peptidoglycan or otherwise, they'll probably just clog the channel instead of stretching.  Still amazing.
You can read more about it at Lab Manager here.

Amine‐reactive neutron‐encoded labelsfor highly‐plexed proteomic quantitation

In press at MCP right now, is a new paper from Josh Coon's lab at the University of Wisconsin.  In it they describe the ability to do 12-plex labeled quan with neutron encoded labels.

12-plex.  The catch, is that the labels differ by 12mDa.  So it's going to take some serious resolution to separate them, so it won't be something you could use on a Q-TOF.  You'll probably even have trouble on most Orbitraps as ultra-resolution appears necessary to pull these labels apart effectively.  In the paper, they demonstrate the very clean separation of these labels at 240,000 and 480,000 resolution.  I'm doing the math in my sleepy head in an airplane, but I think that 60,000 resolution should, in theory, resolve peaks this close together, but that is still a pretty slow scan for most instruments.

So, an awesome new labeling technology, it just might not be accessible for all of us right now.
The article is available here.

Novel chemotherapy topic discovered through quantitative proteomics!

A common criticism of Proteomics is the lack of true chemotherapeutics that have resulted from discovery studies.  (I hope I don't have to say that I disagree strongly with this point).  But I do hear it out there.

A new paper from Henrik Johannson et. al., out of the Karolinksa Institute (Sweden) and just released in Nature Communications describes a new chemotherapeutic target, retinoic acid receptor alpha (RARA) that was discovered almost solely by quantitative proteomics.  You can read the paper, open access, here.

Take that, naysayers!

In-solution digestion -- the ultimate analysis

When looking at digestion protocols, have you ever stopped to wonder things like "is it 15mM because it was optimized, or is it 15mM because it worked okay and no one has changed it?"

If you have, do I ever have the paper for you!  In press at MCP right now is this paper, easily the most comprehensive study of in-solution digestion ever performed.  They didn't cut any corners here.  Thorough, thorough, thorough.

Later, when we wonder why we were using a certain concentration in a protocol, I hope everyone will just refer back to this paper (for in-solution) and others where there was no guesswork.

Highly recommended for anyone doing in-solution digestion, or who has done it in the past and found low efficiency.

Surgical knife coupled with mass spectrometry for cancer profile detection

Photo:  Sang Tan, AP, from USA Today article

USA Today is a strange place to get news about mass spectrometry devices, but this one is novel enough for the mainstream media.  Apparently, it is common practice to cauterize tissue after cancer removal.

The new "smart knife" is connected to a small MS device that is loaded with cancer profiles.  The smoke arising from the cauterized tissue is then compared to these profiles to determine whether the tissue is more like normal cells or the cancerous ones.

There isn't much more info than this in the USA Today article, but there is loads of info in the original article at Science Translational Medicine.

Optimized nonlinear gradients for RPLC-MS

Do you always wonder about which gradient to be used for LC-MS analysis? Although, there is no such thing as perfect LC gradient as each and every sample is somewhat different from the other and one particular gradient will not be suitable for every sample. However, users in the field, usually stick to one gradient that worked best for the most part.

This study recently published in Analytical Chemistry by Luminita Moruz et al., deals with optimizing the gradient in a non-linear way (article). In brief, one needs to analyze the sample with the linear gradient method and then use the gradient and peptide id - retention time information to get the non-linear gradient. The output is a text file containing an optimized non-linear gradient method. Now, this is based on the peptide id and their retention times that originally came from your linear gradient method. You can see the change in gradient in following image post-optimization. I obtained this by using the python script they provided on GoogleCode. I will have some data on this in a few days. In the article, they have shown that the peptides exclusively identified by nonlinear gradient method have somewhat different properties than with gradient method. So, you get two things, one is that even spread of peptides across the LC gradient and more coverage of peptides/proteins.

The python code is available at Google Code here.

SMP Proteomics Conference August 12-16th

Need some sun with your Proteomics conference?
Then maybe you should check out the SMP Conference in Cancun August 13-16.

Invited speakers include:

Laura Beretta (University of Texas)

Al Burlingame (University of California-SF)

Juan Calvete (CSIC-Spain)

Catherine E. Costello (Boston University)

Marcos Eberlin (Universidade de Campinas- Brasil)

Sergio Encarnación Guevara (Center of Genomic Sciences)

Emer Ferro (Universidade de Sao Paulo-Brasil)

Benjamin García (University of Pennsylvania)

Donald Hunt (University of Virginia)

Hanno Langen (Translational Research Sciences -La Roche, Switzerland)

Cesar López-Camarillo (Universidad de la Ciudad de México)

Luis Mendoza (Institute of System Biology-USA)

Mathias Selbach (The Max Delbrück Center for Molecular Medicine, Berlin)

Andrej Shevchenko (Max Planck Institute-Germany)

Gary Siuzdak (Scripps Center for Metabolomics and Mass Spectrometry)

Robert Winkler (CINVESTAV-Irapuato, México)

Extract phosphopeptides from Proteome Discoverer output

Have you ever wanted to just pull the phosphopeptides out of your Proteome Discoverer output data?  Well, I did.  During my second postdoc, I primarily did global quantitative (SILAC) phosphoproteomics.  In order to extract the data, my friend Dr. Trevor Glaros (Proteomics Scientist, now at USAMRIID) and a colleague at Virginia Tech developed a macro that could pull out these valuable mods.  This was a big step forward for my research, as well as a valuable lesson in my life, that sometimes it is better to know people than it is to try to do everything yourself.

Anyway, I'm going to share this valuable Excel sheet with everyone.  You can access it here:  http://orsburn-useful.bytebuckets.com/

As a side note, I found this lesson so valuable (and this formula so useful) that I had it tattooed on my leg a while back so that I'd never forget it.

PD-nodes.org updates, nodes available for PD 1.4

More quietly released updates at PD-nodes.org!

All nodes are now available for download into PD 1.4 (except spectrum merger node, which is also available as a stand-alone)

A poster and a thorough presentation on the workings of MSAmanda is now available through the site as well.

A large resource of synthetic peptides and their phosphopeptide counterparts

Dr. Bernhard Kuster's group from Technical University of Munich in collaboration with Dr. Matthias Mann, Dr. Albert Heck and Dr. Shabaz Mohammed, has generated a large resource of >100,000 synthetic peptides and their phosphopeptide counterparts.
In a recently published article in Nature Biotechnology, they have described in detail about it. This has many applications and some of them have been nicely put in the article such as HCD v/s ETD fragmentation of peptides, comparison of Mascot delta score, PTM score and PhosphoRS and many more. This will be really helpful to the scientific community.

Dirty genomics and proteomics

The term "dirty genomics" is a new one to me as of this morning, but it is good name for something that is coming up all the time now:  how do we use this wealth of high throughput genomics data that keeps pouring down the pipeline?

Many of these sequences are incomplete and virtually all are annotated by algorithm only.  From what I've seen, these algorithms struggle with fitting the sequences into a specific template (like FASTA).  But they are undoubtedly very valuable.

For many organisms, these incomplete sequences are all that are available.  On top of that, I have heard from a number of researchers that supplementing the standard human FASTA databases with incomplete sequencing data from that particular cell line strain (or person) adds significant value to the proteomics data set.

The published information seems a bit limited at this point and appears to be limited to the microorganisms, but there is some out there, including this 2009 PlosOne article where they had success using an incomplete sequence.

I'm not going to introduce a solid opinion here, I need more data, but I really wanted to kick out this dirty genomics idea and the thought that this may be a resource we can use to improve our results.

Biomerieux and Ultimate Labs Partner on Development of MALDI Microbe ID Platform

I have been continually fascinated by the technologies that use MS-based assays to rapidly and effectively identify micro-organisms.  While these are commonplace as backup systems in most big hospitals and clinical facilities, the technologies are (as far as I have seen) almost entirely GC-MS systems, where the GC is really the primary component.  I hear rumblings about MALDI systems and have seen successful small scale experiements in the literature, but this new system in development by the partnership between these two companies seems primed to move to center stage.  The system is currently in review for certification by the USDA.  For more information, you can follow this link to Genome Web (requires that you are registered to read, but it is free) or at the company home page.

New proteomics jobs added to the jobs page!

I added 9 new positions today to the job page.  One of which is an opening with my team:  A Senior Applications Scientist in Small molecule high resolution mass spectrometry.  This is reserved for someone who is really good and wants to join a top notch team of scientists.  Check it and these other opportunities out here.

Maximum peptide identifications in data dependent acquisition experiments

Do you love your standard data dependent experiment, but hate the low level of reproducibility between technical replicates?  If you answered "Yes!" then you really should check out this new paper out of Christine Wu's lab at Pitt.

In this study, the authors take a look at the weakness in the standard DDA experiment between technical replicates.  Instead of relying only on peptides that were positively ID'ed in each run, they examined the XICs of the identified peptides from each individual run.  The high accuracy MS1 and retention time from each identified peptide was compared to those of the other runs, even when sequencing information was not available from the second run.

By running the experiment and interpreting the data in this way, rather than at the level of mutually identified peptide IDs, no loss in peptide IDs are reported.  The level of reproducibility between runs, however, increases dramatically.

As you'd expect from the names on this paper, the work is solid, thorough, and innovative.  I strongly recommend that everyone doing shotgun proteomics take a look at this, because it isn't often that we get better results without compromising something (or spending more money!)

Proteomics tools freely available at the Google Code !!

Most of the times researchers wonder as how to do a certain data analysis and then look for correct tool to get their work done. It might be some formula within MS Excel. When I get into that situation, I try to find the relevant tool at the Google Code.

Google code is a great place to find some of the open source tools and packages for various fields. I came across one such tool few months ago, PeptideShaker - It is useful to see search results from Mascot, OMSSA and X!Tandem. Other tools include Peptizer, Compomics,  dbtoolkit etc. Its always helpful to have some freely-available tools by your side.

Proteomic Applications in Cancer Detection and Discovery

This month's SeparationsNOW newsletter highlights Tim Veenstra's new book shown above.  It can be ordered now directly from Wiley here, or you can purchase the it for your Kindle here.

GeneWiki for iOS -- Protein and gene ontology information on the go.

Do you need access to the most up-to-date annotations for human and mouse genes and proteins when you are away from your lab?  Do you have lots of patience and the time to keep touching your iOS screen so it doesn't go to sleep and reset an App?  Then do I ever have the $.99 App for you!

Getting the App to start requires a little dedication, as it first has to download and compile it's databases before you can use them.  On a hotel wifi network the downloading can take quite a while.  The compiling takes some time wherever you are (roughly 15 minutes).  During the download process if you switch to a different App or allow your iOS device to go to sleep, the process resets completely and you must begin again.  On the third time your iPAD has gone to sleep at 50% or more of the way through the process, you might find yourself tempted to delete the App.

Though if you are someone with more spare time (and patience) than I currently possess you'll reap the rewards of the dollar and half hour of your time you spent just touching your iPAD.  I don't know if the program is available for Android, perhaps the multitasking capabilities of those modern devices will allow the program to complete it's machinations in the background.

You can find more information about what GeneWiki supposedly does here.

Jobs in Proteomics

Do you have a job posting for a proteomics position?  Technical, postdoctoral, or otherwise?  Email me:  orsburn@vt.edu and we'll post it here.  Click here to go to the new job page.

Shared Proteomics Interviews -- Mike MacCoss and Josh Coon

SharedProteomics has made a habit recently of interviewing some of the top innovators in our field. Their collection of interviews is now something that can be reached directly from the home page.  The most recent additions have been Josh Coon and Mike MacCoss.  You can read about these and other names we all know here.

SympatiQCo

Automated quality control analysis of accumulated RAW data?  Am I sleeping?  After years of dreaming about quality control being taken seriously in Proteomics, the last year has been a revelation!  People all over the place are talking about great things like the PRTC (Peptide retention time calibration) standards from Pierce and new software is whispered about all over the place.

But this one is real and from (who else?) our friends at IMP and the Austrian Academy of Sciences.  SympatiQCo is the product of Michael Mazanek and is meant to be a server based system to automatically QC your LC-MS runs.  It looks at peak shape, retention time and can even perform Mascot test searches if a server is available.

Definitely go here to find out more!

Is ProteoIQ going the way of Elucidator?

Back in May, ProteoIQ suddenly became the property of Premier Biosoft.  I generally let news like this slide by, mostly because I want to stay out of it.  This week, however, I learned from several customers that the changes haven't been entirely favorable.  Both customer service and technical support seem to have taken a hit in the move.  Hopefully Premier will get it together and not let this nice software package go the way of the now legendary Elucidator suite!

R for Proteomics

Nothing makes me happier when rigorous statistical analyses are applied to genomics and proteomics data sets.  That is why I had to do a serious double-take when I was informed about R for Proteomics.

If you aren't aware, R is an open source statistics platform that is supported by a large community of programmers and scientists.  It is used in everything from population dynamics to microarray studies.  The real flexibility in R is gained by the fact that small programs can be relatively easily integrated into R as it has no problem with programs written within its own sub-language or C or C++.  The real power comes from the fact that the community actually develops new programs for it and supports them.

R for Proteomics is a smaller sub-branch that is beginning to gain some steam.  A new review by Laurent Gatto and Andy Christoforou is currently in press at Elsevier that highlights some of these programs and their capabilities.  And there are many, and all of them adhere strongly to the core philosophy of R, that robust statistics come first.

For more information, follow this link.

Phosphoproteomics of Type 4 Pilus construction

When I was in graduate school, I heard a LOT about the construction of Type 4 pili.  This was because I worked down the hall from one of the top Myxococcus labs in the world.
So when I was digging through the papers in press at MCP and I saw this new one that deals with the phosphoproteomics of Type 4 pili construction, it caught my eye.

The subject of the paper is a thermophilic bacteria that uses Type 4 pili to get around.  The authors use titanium dioxide to enrich for phosphopeptides from the bacteria while it is moving under different high temperature conditions.  An Orbitrap XL does the grunt work and helps identify the key phosphorylation sites for this strange and interesting organism.

The paper is the result of work at several top universities in Taiwan and good  read for any of my microbiologist friends out there!

Book recommendation

This is somewhat on-topic, but also kind of a distraction.  I've nearly finished this book and I highly recommend it.  We are seeing this term "Big Data" everywhere.  From the covers of the best academic journals to those of magazines in the airport bookstores, we're surrounded by this term.  This book provides a really good explanation of what this really means.  While the examples they use are for everyone, like how the Amazon recommendation engine works and how Google tracks the spread of the flu, I'm sure you'll find yourself thinking of ways to apply these concepts to our field.

Big day for the blog!

This is a big day for the blog.  The staff has doubled.  Santosh Renuse of the Institute of Bioinformatics in Bangalore has joined in my efforts here and will be a contributing author.  I hope these changes will help to move the blog into more of what the title says it is:  an outlet for News in Proteomics.  Sometimes I feel that I get too caught up in specific applications to keep a good eye on the major new advances in our field.  With double the contributors I think we'll have a better chance of making this what I've always hoped it could be.  I expect some exciting changes.

Chorus Project -- My favorite thing from ASMS 2013!

The Orbitrap Fusion is awesome.  The releasing of the most sensitive triple quads of all time is a great advance for everyone.

My personal favorite thing to come out of ASMS 2013, however,  is the Chorus Project.  I`m going to rant for a second on why that is.

A few years ago this extremely prestigious genomics researcher at Duke got busted for faking genomics data.  Not small time faking, BIG TIME faking.  He had convinced people that his microarrays and analysis techniques were so accurate that he could do cDNA microarrays on patients and tell what chemotherapy would work best for their cancer.  And people were doing it!

Of course, it didn`t start there.  It started with several completely fabricated studies that had made it to the very very best journals.  And they made it because the array data was just too big and dense for anyone to manually review.

Almost all genomics journals require the dropping of genomics data into repositories for just this reason.  We had the same thing in proteomics.  You couldn`t publish in MCP until your data was up on Tranche.  Unfortunately, our data kept getting better and that meant bigger, and now Tranche can not handle even a fraction of the proteomic data being published.

We currently have no oversight.  We can publish anything, and there have been a few papers recently (no hints on which, what, or who) that have had findings that are a little beyond belief.  We are currently at a point where our field can be seriously hurt by liars and charlatans and there has been no solution in site.

Enter the Chorus Project.  Proteomics processing and data sharing viewing through the Cloud.  Cloud sized storage that can handle even the next generation platform data files  AND proceess them (Comet processing on the cloud!) AND make the data easily viewable from any web connection.  You see the data without having to download all 700 GB of it.  It is incredible.

Disclaimer:  It is still in Beta version, but it is coming.  I have been on it (Beta tester! Woo!), and I have seen it run.  I dream of a day when I can read a paper with amazing new findings that push the boundaries of my belief and logging onto Chorus and seeing it right there and being able to believe it!

I know you want to know more about Chorus and the great minds who are bringing it to us.  So go to https://chorusproject.org and check it out!

Top10? Top20? Top30? What experiment do you run on the Fusion?

I know I touched on the Fusion a little, right before I left.  But I just want to touch on one of the coolest features of the instrument, called Dynamic Scan Management (DSM).

If you've been following my blog, you know that I've spent considerable time providing my recommendations for optimizing LC-MS parameters with the appropriate number of MS/MS events.

DSM negates part of this.  Due to the architecture of the Fusion, it can, like the hybrids collect MS/MS scans while the MS1 is running.  Unlike the hybrids, however, the Fusion will automatically perform the maximum number of MS/MS scans that can be during the MS1 fill and acquisition time.  This way you get the highest possible number of MS1 and MS/MS scans for every experiment at your particular resolution.  And, BTW, thanks to HCD ion trap scans, the Fusion can get as many as 20 scans/ second!

I have spent a week working with 2 Orbitrap Fusion models at the San Jose demo labs and I can promise you that this thing is a big step forward from the hybrid systems, definitely in part due to DSM.  On top of that, not having to optimize your MS/MS to your LC and sample complexity really takes some of the guesswork out of setting up each experiment.  I hope to have some data to show later on samples ran on an Orbitrap Velos Pro followed by a run on a Fusion in the next few days.

Bloggers needed!

I came to a sudden realization yesterday.  I don't think I can do this alone anymore.  I really want to keep this going, but sometimes I'm going to take 3 weeks of vacation starting in the middle of ASMS and am going to get behind on everything.
So, I decided that I am willing to start opening up this blog to other writers.  Over the years I've had a request or two, but I had more spare time and wanted complete control over this forum.  If you are interested in contributing, drop me a line at: orsburn@vt.edu.  Let's keep this thing rolling.
Even if no one else jumps on, there is just SO much for me to write about right now.  Expect a lot this week (once my new laptop arrives tomorrow, that is!  I'm also having some technical difficulties, btw!  Double excuse!)