Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Correspondence
  • Published:

moFF: a robust and automated approach to extract peptide ion intensities

Nature Methods volume 13, pages 964–966 (2016)Cite this article

Subjects

To the Editor:

We present modest Feature Finder (moFF; https://github.com/compomics/moff): a fast, robust, operating-system-independent relative quantification algorithm that scales to enable quantitative analysis of very large mass-spectrometry-based proteomics data sets.

This is a preview of subscription content, access via your institution

Relevant articles

Open Access articles citing this article.

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Figure 1: moFF performance compared to MaxQuant.

References

  1. Vaudel, M., Sickmann, A. & Martens, L. Proteomics 10, 650–670 (2010).

    Article  CAS  Google Scholar 

  2. Liu, N.Q. et al. J. Proteome Res. 12, 4627–4641 (2013).

    Article  CAS  Google Scholar 

  3. Sandin, M., Teleman, J., Malmström, J. & Levander, F. Biochim. Biophys. Acta 1844 (1 Pt A), 29–41 (2014).

    Article  CAS  Google Scholar 

  4. Cox, J. & Mann, M. Nat. Biotechnol. 26, 1367–1372 (2008).

    Article  CAS  Google Scholar 

  5. Vizcaíno, J.A. et al. Nat. Biotechnol. 32, 223–226 (2014).

    Article  Google Scholar 

  6. Vaudel, M. et al. Proteomics 16, 214–225 (2016).

    Article  CAS  Google Scholar 

  7. Vaudel, M. et al. Nat. Biotechnol. 33, 22–24 (2015).

    Article  CAS  Google Scholar 

  8. Vaudel, M., Barsnes, H., Berven, F.S., Sickmann, A. & Martens, L. Proteomics 11, 996–999 (2011).

    Article  CAS  Google Scholar 

  9. Argentini, A. et al. Protocol Exchange http://dx.doi.org/10.1038/protex.2016.085 (2016).

  10. Paulovich, A.G. et al. Mol. Cell. Proteomics 9, 242–254 (2010).

    Article  CAS  Google Scholar 

  11. Staes, A. et al. Anal. Chem. 85, 11054–11060 (2013).

    Article  CAS  Google Scholar 

Download references

Acknowledgements

This research was supported by the Ghent University Multidisciplinary Research Partnership “Bioinformatics: from nucleotides to networks” (01MR0410) to L.J.E.G., L.C. and L.M., Ghent University grant BOF12/GOA/014 to N.H. and L.M., the IWT SBO grant “INSPECTOR” (120025) to A.A. and L.M., and IWT SBO grant “Differential proteomics at peptide, protein and module level” (141573) to L.J.E.G.

Author information

Authors and Affiliations

  1. Department of Medical Protein Research, Ghent, Belgium

    Andrea Argentini, Ludger J E Goeminne, Kenneth Verheggen, Niels Hulstaert, An Staes & Lennart Martens

  2. Department of Biochemistry, Ghent University, Ghent, Belgium

    Andrea Argentini, Ludger J E Goeminne, Kenneth Verheggen, Niels Hulstaert, An Staes & Lennart Martens

  3. Bioinformatics Institute Ghent, Ghent University, Ghent, Belgium

    Andrea Argentini, Ludger J E Goeminne, Kenneth Verheggen, Niels Hulstaert, Lieven Clement & Lennart Martens

  4. Department of Applied Mathematics, Computer Science and Statistics, Ghent University, Ghent, Belgium

    Ludger J E Goeminne & Lieven Clement

  5. Department of Plant Systems Biology, VIB, Ghent University, Zwijnaarde, Belgium

    Ludger J E Goeminne

Authors
  1. Andrea Argentini
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Ludger J E Goeminne
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Kenneth Verheggen
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Niels Hulstaert
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. An Staes
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Lieven Clement
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Lennart Martens
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Lennart Martens.

Ethics declarations

Competing interests

The authors declare no competing financial interests.

Integrated supplementary information

Supplementary Figure 1 moFF workflow overview

The optional match-between-runs module takes as input a list of features of interest for a given run, and then matches the corresponding features in other runs. Upon matching, all peptides (either identified directly, or indirectly by matching between runs) are processed by the Apex intensity module to extract feature intensity and associated quality measures. The results are then written to tab-delimited output files

Supplementary Figure 2 Overview of the apex intensity module

Around the RT of the input feature, a time window is constructed which results in a local XIC. The peak apex is then located in this local XIC. The log_L_R metric measures the skewness of the peak around the RT of the obtained apex point. The SNR metric provides the ratio of peak height to noise, where the noise value is set as the lowest intensity value in the local XIC

Supplementary Figure 3 Boxplots of the shape measure (log_L_R), SNR, and RT difference

Boxplots of shape measure (log_L_R), SNR, and RT difference between input peak RT and peak apex RT as identified by moFF, across all CPTAC study 6 runs. The majority of peaks across these 45 runs display a symmetrical shape (log_L_R close to 0) and consistently high SNRs. Replicates within each of the labs (designated by a set of three consecutive numbers: 1, 2 and 3 for lab 1; 4, 5 and 6 for lab 2; 7, 8 and 9 for lab 3) show quite consistent median RT difference, while more variation can be seen between labs. The samples (designated by capital letter A, B, C, D, E) have less of an effect, with different samples showing consistent median RT differences for a given lab.

Supplementary Figure 4 Pearson's Correlation of moFF and MaxQuant intensities and quality metrics on the human dataset

Pearson's Correlation of moFF and MaxQuant intensities, boxplots of shape measure (log_L_R), SNR, and RT difference between input peak RT and peak apex RT as identified by moFF in the 20 fractions of the human sample. On this data set too, moFF and MaxQuant show extremely high Pearson's correlation coefficients (between 0.93 and 0.97). The peak shapes show a more left skewed shape for these 20 fractions. The SNR values also show higher median values than observed for the CPTAC samples. Similar to the CPTAC study 6 data, moFF finds the apex peak within 0.10 minutes of the expected RT on average, which is only a third of the window size allowed to moFF

Supplementary Figure 5 Scatter plots of the moFF and MaxQuant intensities for all the 15 runs related to CPTAC lab 1

The color from blue to red shows the distance of each point to the linear regression model that fits the data. Red points are the extreme outliers considered in Supplementary Table 1.

Supplementary Figure 6 Scatter plots of the moFF and MaxQuant intensities for all the 15 runs related to CPTAC lab 2

The color from blue to red shows the distance of each point to the linear regression model that fits the data. Red points are the extreme outliers considered in Supplementary Table 1.

Supplementary Figure 7 Scatter plots of the moFF and MaxQuant intensities for all the 15 runs related to CPTAC lab 3

The color from blue to red shows the distance of each point to the linear regression model that fits the data. Red points are the extreme outliers considered in Supplementary Table 1.

Supplementary Figure 8 Evaluation of three match-between-runs options in terms of standard deviations of log2-transformed MS1 intensities of shared peptides in moFF and MaxQuant.

Following match-between-runs are considered: a) outlier filtered and unweighted b) no outlier filtering and unweighted and c) no outlier filtering and weighted. In each plot (right and left part) we also show the effect of the filtering peaks with a SNR at or below the 5th percentile of all the MS2 SNR values. For each lab, peptides were subdivided in three groups: those peptides where match-betweenruns was employed (mbr), those peptides whose features were solely matched by MS2 identifications (ms2), and the union of all peptides (all). The use of outlier filtering and weighting scheme (see Supplementary Methods) does not have an effect on the MS1 intensities of the moFF matched peptides.

Supplementary Figure 9 The RT difference between the predicted RT by moFF and the MaxQuant RT for the matched peptides in case of 15 replicates for each CPTAC lab

For each lab we evaluated four different options (outlier filtering on/off, and weighted/unweighted terms) that are provided in the matchbetween-runs module of moFF: i) outlier filter and weighted term; ii) outlier filtering and unweighted term; iii) no outlier filtering and weighted term; and iv) without outlier filtering and unweighted term. We notice that filtering the outliers in the training phase of the models show a more visible reduction of the RT error for CPTAC lab 1. For all four options, the median RT difference is low (below 0.2 minute).

Supplementary Figure 10 Comparison of the standard deviation of the MS1 intensities for all features, ms2 identified features, and match-between-run features for moFF and MaxQuant, related to CPTAC lab 1, using 3 replicates for each sample

The left and right plots refer to the unfiltered and filtered matched features found by moFF, respectively (filtering was applied using the same criteria used for the 15 replicates analysis used in Supplementary Figure 2).

Supplementary Figure 11 Comparison of the standard deviation of the MS1 intensities for all features, ms2 identified features, and match-between-run features for moFF and MaxQuant, related to CPTAC lab 2, using 3 replicates for each sample

Left and right panels present unfiltered and filtered matched features, respectively.

Supplementary Figure 12 Comparison of the standard deviation of the MS1 intensities for all features, ms2 identified features, and match-between-run features for moFF and MaxQuant, related to CPTAC lab 3, using 3 replicates for each sample

Left and right panels present unfiltered and filtered matched features, respectively.

Supplementary Figure 13 RT difference for the predicted RT of moFF and the MaxQuant RT for the matched peptides for CPTAC lab 1 in case of 3 replicates per sample

For each sample we have evaluated four different options present in the match-between-runs module of moFF: i) outlier filter and weighted term; ii) outlier filtering and unweighted term; iii) no outlier filtering and weighted term; iv) no outlier filtering and unweighted term

Supplementary Figure 14 RT difference for the predicted RT of moFF and the MaxQuant RT for the matched peptides for CPTAC lab 2 in case of 3 replicates per sample.

For each sample we have evaluated four different options present in the match-between-runs module of moFF: i) outlier filter and weighted term; ii) outlier filtering and unweighted term; iii) no outlier filtering and weighted term; iv) no outlier filtering and unweighted term.

Supplementary Figure 15 RT difference for the predicted RT of moFF and the MaxQuant RT for the matched peptides for CPTAC lab 3 in case of 3 replicates per sample

For each sample we have evaluated four different options present in the match-between-runs module of moFF: i) outlier filter and weighted term; ii) outlier filtering and unweighted term; iii) no outlier filtering and weighted term; iv) no outlier filtering and unweighted term

Supplementary information

Supplementary Text and Figures

Supplementary Figures 1–15, Supplementary Tables 1–3, Supplementary Notes 1–3 and Supplementary Methods. (PDF 2562 kb)

Source data

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Argentini, A., Goeminne, L., Verheggen, K. et al. moFF: a robust and automated approach to extract peptide ion intensities. Nat Methods 13, 964–966 (2016). https://doi.org/10.1038/nmeth.4075

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/nmeth.4075

This article is cited by

Search

Advanced search

Quick links

Nature Briefing: Translational Research

Sign up for the Nature Briefing: Translational Research newsletter — top stories in biotechnology, drug discovery and pharma.

Get what matters in translational research, free to your inbox weekly. Sign up for Nature Briefing: Translational Research