Looking ahead to some 2012 events

The start of 2012 has been a very busy time for us in terms of organising our attendance at conferences which are taking place all over the world. I have been making final arrangements for events being held over the next couple of months, and booking up meetings for later in 2012.

Our first conference in Europe this year is the Joint Conference of Polish MS Society and German MS Society  4th – 7th March. Usually the German MS Society (DGMS) have their own annual meeting, but this year they are having a joint meeting with their Polish counterparts in Poznan, Poland.  We will be represented by Leo Pollack and Agnès Corbin, so please visit us to see the latest developments in the Progenesis software range, including a preview of the next Progenesis CoMet release.

 

Following on from this, our next conference is across the Atlantic in Orlando, Florida. ABRF 2012 – Learning from Biomolecules, The Technology Behind the Story is being held at Disney’s Contemporary Resort, March 17th – 20th. The focus of the meeting is the cutting edge technologies that drive scientific research, and how to implement these techniques in core facilities.

One of the highlights of the meeting for us is the T5 Tutorial Session – Overcoming Problems with Experiment Design for Quantitative Proteomics. It is being held on  Tuesday 20th March, 4:30pm – 5:45pm, and will be hosted by my colleague, Paddy Lavery.  There are presentations from Nonlinear’s founder and CEO, Will Dracup, Julian Whitelegge Ph.D from UCLA and the session closes with a presentation by Juan Chavez, from the Bruce Lab, University of Washington.

The workshop aims to highlight the challenges of experiment design and the statistical analysis of ‘omics data. We hope that the people who attend the workshop will learn more about how to do reliable and reproducible proteomics experiments.

Please let us know if you would like more information about this, or any of the other conferences we are attending in early 2012.

Looking forward to the rest of the year, we recently confirmed our attendance at the following conferences:

103rd AOCS Annual Meeting & Expo

29th April – 2nd May 2012

Long Beach, California, US

More information

60th ASMS Conference on Mass Spectrometry and Allied Topics

20th – 24th May 2012

Vancouver, BC, Canada

More information

HUPO 11th Annual World Congress

9th – 13th September, 2012

Boston, MA, US

More information

This is just a selection of the conferences we will attend over the next few months, you can see the full, up-to-date list on our website, or keep in touch with us on the blog.

Thanks

How can you analyse a mixture of technical and biological replicates in your proteomics experiment?

Guidelines for submitting manuscripts for publication require you to consider both technical and biological replicates. A typical example can be seen in the notes to authors from PROTEOMICS and PROTEOMICS – Clinical Applications (Wiley-VCH).

To satisfy this, many people run both biological and technical replicates as part of the same experiment. This adds unnecessary complexity to your data analysis and, unless handled correctly, you risk pseudo-replication.

As an alternative, you can consider setting up experiments that comprise multiple technical replicates of one biological sample. These are then used as a measure of the technical variation of your system and approach. You then compare any changes observed between biological replicates against the background of technical variation to report the reliability of the biological differences.

In this post, we’ll consider an experiment where both types of replicate are run together, since this is often what occurs due to limitations in time or resources. Defining your experimental groups as containing both technical and biological replicates would introduce bias and error, making your results invalid.

So, how do you analyse both technical and biological replicates to give statistically valid results?

Analyse technical and biological replicates within a single experiment

The approach to analysing a mixture of technical and biological replicates with Progenesis is like reproducing the whole experiment, but doing this easily in silico, and comparing two or more sets of results to find significant features common to both. This is true whether you use Progenesis SameSpots, Progenesis LC-MS or Progenesis CoMet.

Summary: Progenesis products can be used to analyse experiments that contain a mixture of technical and biological replicates in an organised, statistically valid way. The result is like reproducing the whole experiment and comparing two or more sets of results to find significant common to both, the intersection in the Venn diagram above.

How do you do this in Progenesis?

The key is in the unique way Progenesis products align and co-detect runs or images, which means the same features are detected and measured identically on every run or gel image. This allows you to easily compare different views of the same data, comparing features across any runs/gels that are set up in different experimental groups. We can demonstrate this with an example of a study comparing Control versus Treated subjects containing a mixture of technical replicates, A and B, as well as biological replicates 1, 2 and 3.

• Figure 1: Set up all files in one experiment, align and detect then set up multiple experiment designs within the same experiment. One experiment design should contain all files together and the other experiment designs only contain the technical replicates of runs or images from each condition. In our example below, we have created an experiment design with all technical and biological replicates from the control subjects versus the same for Treated subjects. We also generate an experiment design comparing technical replicate A’s for Control versus Treated and another comparing technical replicate B’s for Control versus Treated.

Figure 1: Set up all files in one experiment, align and detect then set-up multiple experiment designs in the same experiment.

• Figure 2: Once you step into the View Results section, you need to: 1. Select the experiment design based on containing technical replicates, A or B in our case, from the drop down option (highlighted in red below).  2. Apply tags to features considered significant based on your chosen criteria within each experiment design. Then apply tag filters to view the discoveries common to both replicate analyses, as well as those that are common to only one replicate set.

Figure 2: 1.Tag features with significant criteria e.g. p-value <0.05, fold-change >2 of Control versus Treated in both experiment designs i.e. All technical replicate A’s for control vs. Treated and all technical replicate B’s for Control vs. Treated. 2. Apply tag filters to only list the features with significant criteria common to both experiment designs.

• Figure 3: By using the tags and applying tag filters, you can identify the specific features that are common to each analysis. From here, you check the number matches your FDR and these features can be ignored in any necessary validation work.

Figure 3: The final analysis results from the example of Control versus Treated samples made up of a mixture of technical and biological replicates. Here we could take the 68 common significant features and select those for further study.

Want to learn more?

This is a very simple overview and example, so you may wish to see a demonstration of this with your own data. Get in touch, tell us if you are interested in seeing the workflow for 2D gel analysis or label-free LC-MS data analysis and we can arrange to set this up with you. Hopefully we can also show you what else Progenesis SameSpots, Progenesis LC-MS or Progenesis CoMet can do to help your proteomics or metabolomics analysis.

Happy Holidays from all at Nonlinear

We have reached that festive time of year when a lot of people, including us here in the UK, are about to take some time off to spend with family and friends. And that includes the authors of this blog.

We are already looking forward to next year, which promises some exciting new product releases and much more. There are lots of plans in the pipeline for 2012, but at  the moment, we can reflect on a pretty good 2011!

Highlights included:

Progenesis CoMet v1.0  – a new product developed to measure the relative abundances of metabolites.

Progenesis LC-MS v4.0 – with an  extended workflow for the quantification and identification of proteins which includes multivariate statistics.

Progenesis SameSpots v4.5 – allows users to link protein search results to spot data and put their 2D analysis into biological context.

We would all like to wish you all the best for the season, and we’ll be back on the blog with more interesting news and updates in 2012.

Happy holidays!

Best wishes,

Beth, Paddy, Mal, Wilka, Jules, Agnès, Mark and Paul

Free download: improved cross-platform support in CoMet and LC-MS

In recent months, we’ve broadened our support for data file formats in the latest versions of both Progenesis LC-MS and Progenesis CoMet. Thanks to the plug-in architecture that we use in Progenesis, you can get these improvements today, free of charge.

mzML and netCDF support for Progenesis CoMet

Earlier this month, we introduced support for both the netCDF and mzML run file formats in Progenesis CoMet. Consequently, if your lab uses a Shimadzu ion trap or an instrument from AB Sciex, you too can now try out Progenesis CoMet on your own data.

• For Shimadzu instruments, export your runs as netCDF using the relevant data export module
• For ABSciex machines, convert your WIFF files to mzML using the MS Data Converter

Other improvements

Four other plugins have also been updated to provide enhanced support:

• Thermo .RAW files (LC-MS | CoMet)
  Improved diagnostics and wider support for files produced on Orbitrap Elite instruments.
• Proteome Discoverer (LC-MS)
  Now supports .xlsx files as well as the older .xls format.
• Waters .Raw folders (LC-MS | CoMet)
  Now includes support for MassLynx v4.1 centroided data.
• Bruker Daltonics .d folders (LC-MS | CoMet)
  Improved diagnostics for runs that fail to import (typically requires CompassXport to be installed).

As you can see, despite already offering a wide range of file format support, we remain committed to continually improving it. Keep an eye on both the blog and the help panel at the right side of your software’s welcome screen for further updates in the future. And if your preferred file format, search engine, or inclusion list format isn’t yet supported, get in touch to let us know. Who knows – maybe your update could be listed here next time?

3 ways to routinely QC your 2D analysis

If you visit our blog regularly, you will have noticed a recent theme around quality control (QC) for proteomics data and how our analysis solutions can support this. It’s one we have chosen to highlight for many reasons. Fundamentally, as my colleague Beth pointed out in this recent post, “…it’s a case of garbage in, garbage out”. This is especially true for the challenging experiments that proteomics demands. How to address QC in proteomics is becoming a hot topic, with talks at conferences and review articles citing QC as essential for translating quantitative proteomics discoveries into clinically relevant results.

I want to continue the theme and focus on where  QC seems most mature in this respect, the analysis of 2D gels, by highlighting a recent publication.

The example comes from work published by Clémence Bièche et al at the French National Institute for Agricultural Research, as well as other organisations they collaborated with, who use Progenesis SameSpots v4.0. As well as applying QC at various stages in 2D gel analysis the conclusion of their experiments satisfied another hot topic in proteomics; putting results in the context of biological processes.

You can read the paper for full details of the study. Here, I just want to pick out the three key points in the analysis workflow that were used to check and maintain the quality of results they generated:

1. Image QC

Before any image processing the software applies image checks. In this case only those images which passed available QC checks were accepted including ensuring dynamic range was 85-96% and intensity levels of >96%.

2. Principle Components Analysis (PCA)

The experiment compared three different conditions -   0mins, 60mins and 120mins post-HP treatment – to a non-HP control, using a mixture of technical and biological replicates. PCA showed a clear separation of control gels and gels of 0mins. The gels at 0mins were also clearly seen as distinct from the 60 and 120mins post-treatment samples, which were clustered together based on 2D gel measurements. This indicated the gels could be used to define differences in the proteome as it recovered after treatment. Although the gels from 60mins post-treatment and 120mins post-treatment clustered together, the PCA plot showed that 120min samples had more similarity to the control than 60min samples. This indicates the samples measured at the longest time, post-treatment, have proteomes that were recovering enough to resemble the untreated, control samples.

3. p-values, q-values and power calculations

Some thresholds that were applied to these measures ensured only robust statistical differences were used to define the spots that characterised the proteome changing over time. In this case only normalised spots that met the criteria of p <0.05, q <0.05 and power >0.8 were validated and reported as being significantly different. We have FAQs on p-values, q-values and power analysis if you want to learn more.

 

 

These QC measures are automatically generated as part of the main analysis workflow in Progenesis SameSpots, so they provide a complementary series of checks as you head towards a final report of significant spots. As in this publication, the spots of interest can be selected for picking and identification by mass-spec. And with the latest version of SameSpots this protein identification information can be imported and linked to the spots in your experiment. This provides a complete picture based on high-quality results and allows you to put results into biological context.

So why not download Progenesis SameSpots and try these QC measure in the main analysis workflow as well as the SpotCheck workflow to quantify your gel running reproducibility? They help generate data you can publish and rely on.

How to maintain confidence in your sample quality over time

If you’re analysing samples that are stored for weeks, months, or even years, you’ll want to know that those samples aren’t degrading unacceptably. Various factors can contribute to degradation¹, some having quite specific effects while others can cause global changes in the sample. Such factors include:

1. the effect of freeze-thaw cycles
2. the temperature at which the samples are stored
3. the length of time in storage

While various steps can be taken to minimise degradation in your samples, how can you actually quantify the degradation over time? This is where SpotCheck, a workflow in Progenesis SameSpots, can help – even if your analysis doesn’t use 2D gels.

You may have already read about how SpotCheck can help with the training and assessment of new staff in your lab. The same quality control principles can give an objective measurement of how a sample has degraded over time. Assuming, that is, that your analysis protocol has already been verified as giving reproducible results.

Best of all, you can try SpotCheck free of charge. More on that later, but first, how does it work?

How does SpotCheck work?

SpotCheck allows you to compare a 2D gel image to a high-quality set of gels – referred to as a gold standard – run previously from the same sample.

This comparison results in a simple Pass or Fail verdict, based on a quantitative measure of the variation between gold standard and the gel being compared to it.

In the example at the right, the gel being compared to the sample has 86.4% of its spots with volumes that are within 3 standard deviations of the same spot in the gold standard. Both of these thresholds – percentage of spots and number of SDs – are configurable. Setting them in line with the baseline variability in your lab’s methodology allows you to detect variation coming from the sample itself.

Create your own gold standard

Even if you don’t already own Progenesis SameSpots, you can try out the SpotCheck workflow today, on your own data and free of charge. Each download of SameSpots comes with a licence that allows you to analyse six of your own images.

To get started, click here to download Progenesis SameSpots and install it. After downloading, you’ll receive an email that contains your licence code – don’t lose this! Then, using a single, recently collected sample, run a set of 6 technical replicate gels that meet your current QC standards and capture the gel images; these will form your SpotCheck gold standard. You’re then ready to create your gold standard; you can follow the process described in the SpotCheck tutorial, but using your own gel images. The analysis will differ only slightly from the tutorial, as follows:

1. After importing your images, you will need to review their suitability using SameSpots’ 8 quality metrics for 2D gel images. If any images fail these QC checks, simply exclude the failed images from analysis – you won’t have used up any of your 6 licences at this point – and run further technical replicates.
2. Before aligning your gel images (a process that is key to SameSpots’ robust analysis), you’ll be prompted to enter your image licence code – this is the code supplied in the email.

Once your gold standard is created, you will be able to validate as many subsequent aliquots as you wish, without requiring further licence codes. Again, details of the process can be found in the tutorial, but it’s really very simple. Each gel typically takes only 2 or 3 minutes to compare (times may vary depending on your computer’s processing power). So, SpotCheck is simple and quick enough to become an integral and routine part of using long-term stored samples, no matter what type of analysis is being performed subsequently.

Why not try it today?

References

1. David H. Jackson and Rosamonde E. Banks, Banking of clinical samples for proteomic biomarker studies: A consideration of logistical issues with a focus on pre-analytical variation. Proteomics Clin. Appl. 2010, 4, 250–270.

Progenesis SameSpots and 2DE: Developing assays to QC biologic production

Earlier this year, I wrote a blog post about how 2D gel analysis with Progenesis SameSpots can be used to develop reliable ELISA assays for monitoring host-cell protein (HCP) contamination during production of biologics.

An example has been published by Edward Savino and others at Centocor Research and Development Inc. (known as Janssen Biotch Inc. since June 2011) in  BioProcess International^TM . It demonstrates the application of 2D gels compared to Western Blots to measure ELISA antibodyreactivity to HCP contamination in the production of a biologic approved for treatment of rheumatoid arthritis. 

At an early stage of the ELISA assay development, two sets of gels were run to separate the proteins within the host cell line used to generate the biologic product. One set was stained and visualised using Silver Stain; the other set was blotted onto PVDF membranes then probed using a rabbit anti-HCP antibody generated in-house.

Characterisation of anti-host cell protein antibody for reactivity detected on a Western Blot compared to proteins from a host cell lysate detected on a silver stained 2D gel¹. Common spots are highlighted and numbered in red.

Protein spots on the Western Blot were aligned to protein spots detected on the 2D gel using Progenesis SameSpots. The result was 369 spots quantified on the immunoblot compared to 423 spots quantified on the silver stained gel. This showed 87% antibody reactivity to host-cell proteins, which they concluded was of suitable quality for their ELISA assay. The published work goes on to show how the ELISA antibody was used to measure effective HCP clearance at different stages of the production process.

Because each biologic is produced using a specific host cell system and manufacturing process, the principles of developing a unique ELISA assay described here can be applied to other cases.

Try a pilot study today

Download Progenesis SameSpots and you can see how it can help to develop the assays you need to QC biologic production within your own processes.

You get six free image licences, allowing you to perform a pilot study running three 2D gels compared to three Western Blots. If you need to help with the image analysis get in touch and we’ll be happy to run through this with you.

 

 

1. Edward Savino, Bing Hu, Jason Sellers, Andrea Sobjak, Nathan Majewski, Sandra Fenton and Tong-Yuan Yang. Development of an In-House, Process-Specific ELISA for Detecting HCP in a Therapeutic Antibody, Part 1. BioProcess International Vol. 9, No. 3, March 2011 pp38-47.

8 quality metrics for your 2D gel images

It is a widely-accepted truth in scientific research that the quality of your results is dependent on the quality of raw data you analyse. Or to put it more bluntly, it’s a case of garbage in, garbage out.

From the very start of our SameSpots development, we understood the negative effect that poor image quality can have and wanted to do something about it. Our solution was to implement an Image QC step at the start of the SameSpots workflow. It performs an automatic check on all of the images which are imported for analysis, assessing key aspects of the raw image data including:

• Image format
• Image compression
• Level of saturation
• Grey scale
• Dynamic range
• Intensity bit depth level
• % intensity level
• Previous image editing

As you can see, it is quite a comprehensive list, and in the example below, this image has passed all of the image QC checks

SameSpots provides feedback and recommendations so you can refine your image capture.  You are still able to proceed with an analysis in SameSpots if one or more of your gel images fail the image QC process, but generally we would advise against it.

Here are some examples of images which failed the image QC checks.

Saturated image – SameSpots shows areas of saturation on the image in red.

8-bit image – in this image, the pixel data has a bit depth of 8, which isn’t adequate for accurate image analysis (we recommend 16 bit).

Low dynamic range – this refers to the actual range of pixel values used. A low dynamic range can mean less precision in the intensities represented by each pixel. The image intensity histogram illustrates the dynamic range of each image.

Image manipulation tools are included so you can rotate, flip or invert an image that you may have positioned incorrectly on the scanner bed, and an image cropping tool allows you to remove any material from the edge of the gel image which you don’t want to include in your analysis.

You can use the image QC step in Progenesis SameSpots free of charge!

If you download SameSpots, and import your images, they will go through the automatic image QC processes and you will get a report on each image. You don’t need a licence of SameSpots to do this; you can QC your image capture on unlimited numbers of images.

Of course, when you have done the QC on your images, you could use the free image licence code you are sent when you download to do a full analysis on 6 of your images. But if you just want to access the image QC tools then we would encourage you to do that. It will really benefit your 2D / DIGE image analysis even if you use ImageMaster™, DeCyder™, Delta 2D, PDQuest or any other 2D image analysis software.

I hope all of your images get the green light.

Analysing a fractionated mouse serum peptidome using Progenesis LC-MS

As a follow up to my last blog post on analysis of fractionated samples using Progenesis LC-MS, here is a published example. Geun-Cheol Gil, Jim Brennan, Dan Throckmorton, Steven Branda and Gabriela Chirica from the Systems Biology department at Sandia National Laboratories (CA, USA) studied the mouse serum proteome using a fractionation approach¹.

The study used a customised chromatography system for reproducible fractionation. The fractions were  analysed by nanoLC-MS with data analysis using the fractionation workflow in Progenesis LC-MS. I spoke with one of the authors, Dr Geun-Cheol Gil, who said:

“We successfully analysed low molecular weight, low abundant peptides in a complex sample using a reliable fractionation process coupled with analysis by label-free nanoLC-MS. Progenesis LC-MS enabled us to analyse the label-free data from fractionated samples in a rigorous and reliable way and help us to generate a global view of the “peptidome” within mouse serum. Our platform, including this software, and our approach also show that by running simple checks you can minimise the fractions needed to achieve good peptidome coverage and so save instrument time and costs in achieving experimental goals.”

The mouse serum includes components with a wide range of MW and concentration of components. The group wanted to specifically investigate the peptidome (low MW, low abundant peptides in the body fluid). They were able to achieve a comprehensive peptidome identification of 357 proteins, covering a  range of biological functions.

3-D montages of the peptide QSENVGLSSELNR of Testis-specific serine/threonine-protein kinase (accession # TSSK1_Mouse) on Progenesis LC-MS. The peptide was generated by trypsin digestion and ANOVA test was performed for the three sample sets. The p value was 0.06 for the peptide¹. Each peak within the green outline represents a single isotope of the same peptide ion.

Optimising Fractionation

While fractionation offers a route to increase protein and proteome coverage, especially of  low abundance components, it can rapidly increase the number of runs needed per experiment. But due to cost and time demands on analysis, it is desirable to minimise the number of fractions  used. In the published study the final results were based on 3 experiments x 4 fractions x 3 repeats, so a total of 36 LC-MS/MS runs.

This number of fractions and repeats was not arbitrary. The study included analysis to determine how many repeats were needed to provide reliable coverage as well as checks on how well the fractionation had worked. A run of 6 replicates was made and coverage achieved in the first replicate (89%) was compared to coverage in subsequent replicates. It was a step worth taking, since no further increase in proteome coverage was seen beyond 3 replicates. So, the group could be confident in achieving maximum coverage with a relatively low number of replicates.

You can also check to make sure how well your fractionation procedure performs. This was achieved by comparing the distribution of identified proteins between each fraction. This was illustrated in the study by the figure below. Ideally the overlap should be minimal.

Distribution of identified proteins among RAM-RP fractions F1, F2, F3 and F4, generated by elution buffers containing 30, 40, 60 and 80% methanol in 0.1% formic acid, respectively ¹.

The final Review Proteins screen can also display this information directly in Progenesis LC-MS. The table of results illustrates, with the numbered blue squares, which fractions each protein identification is based on.

An example not from the publication.

You can also use the peptide fraction chart to show if you have achieved an even coverage of peptides across all fractions or discover if the same quality of results could be delivered with fewer fractions. In the example below then you could remove at least the first fraction and perhaps optimise the fractionation step to generate equal numbers of peptides or proteins per fraction?

An example not from the publication.

In the words of the authors in this published study:

“Platforms which enable comprehensive, high-throughput and time course analysis of bodily fluids in a cost-effective manner can establish proteomic mass spectrometric survey as the method of choice for personalised diagnostics, monitoring and treatment.”

It is reassuring to see the fractionation workflow in Progenesis LC-MS was considered a key component in this platform and the study provides the first published example of this application. Download Progenesis LC-MS and try analysing some of your fractionated samples for FREE.

 

 

1. G.-C. Gil, J. Brennan, D.J. Throckmorton, S.S. Branda, G.S. Chirica, Automated analysis of mouse serum peptidome using restricted access media and nanoliquid chromatography-tandem mass spectrometry, Journal of Chromatography B (2010), doi:10.1016/j.jchromb.2011.03.028

First look: new features in Progenesis SameSpots v4.5

We have been working on the next release of Progenesis SameSpots and I’m pleased to say that v4.5 is now in its final testing phase and will be released very soon.

So…what’s new?

Clip gallery

For most researchers, the output of their work is the presentation of results in a poster or publication. We wanted to make it easier for SameSpots users to capture images and tables as they work through their analysis, so we developed the Clip Gallery.

You simply right click on the image or table you wish to save and you get the option to add a title and some notes which will help when you are putting together your poster or publication. Images are saved as high resolution .png files (300 dpi) which are ideal for print.

If you are writing your publication or poster on a different computer to the one which has SameSpots installed, you can export all of the clips you have saved to a memory stick.

Protein data import

The workflow has been extended so you can now link database search results back to the protein spot from which they originated. This joins up 2D analysis results with MS-based protein identification so you can more easily put 2D results into biological context. Protein identifications can be imported and viewed in the results tables in SameSpots or exported as a .csv file.

This also allows you to use molecular weight markers to cross check the protein search results and help confirm your protein IDs.

We currently support Mascot and .csv data, but we could look at others if there were demand from our customers. Please contact us to discuss.

Automatic reference gel selection

We want to make the SameSpots workflow as objective as possible, as this helps with the reproducibility of results. Previously, reference gel selection was done manually by the user. Now, SameSpots automatically chooses which gel would make the best reference. It does this by doing a quick check to ascertain which gel image gives the best alignment results.

There is still the option to select your own reference gel if you wish to do so.

And more…

• We have redesigned the software home page to include links to the SameSpots support materials and show the latest news and updates from Nonlinear. The new search facility means you can quickly locate experiments in your list.  If you can’t remember where you saved your experiment, there is a link to open the folder location.
• If you use manual picking, it could sometimes get tricky when the spot numbering on the image to guide your picking was obscuring your view. We have made some improvements to make this clearer, and you also have the option to switch off the spot numbers if you don’t need them.

I hope that some of these features are what you have been waiting for. If you have any questions, or want more information, please get in touch, we’d like to hear from you.