Welcome

to the user-interface of SFINX: the "straightforward filtering index" for cocomplex interactomics data analysis. It enables you to identify the true positive protein interactions buried in your dataset, and this in a fast, convenient and highly accurate way. Press the button, load your data and enjoy the power of SFINX.


Basic data (Example file)

Bait identities (Example file)
Download filtered data as tsv: Complete Output Cut-off Output

Input of example data

Check the 'Use examples' check-box in the panel at the left. All example data should load automatically.

Input of your own data

SFINX requires two files as input: a 'Basic data' file, which is a matrix with your input data, and a 'Bait identities' file, which is a list with your baits of interest. For each of the files, you can indicate whether the file has a header, and how the file is formatted: tab-separated, comma-separated or semicolon-separated.

The 'Basic data' file is a matrix with a column for each project and a row for every protein that was ever detected. The cells of the matrix are populated with the observed peptide counts for the proteins per project. (For an example: see the 'Original data' tab for the analysis of the example data.)

The baits of interest always have to be present as found proteins in your 'Basic data' file. If this is not the case, SFINX will not generate any output for these baits.)

Analysis and download possibilities

Immediately after, you can start discovering the data. At the left, the panel will now also contain a slider to determine the strictness of your analysis (higher values = stricter analysis). At the right, you can start analyzing your data, or you can immediately download the complete list of filtered interactors ('complete output') or the chosen subset ('cut-off output'). The downloaded files are immediately ready for use in most common network visualization or spreadsheet programs.

Before you download the data, you can also go deeper into the results in the tabs above. The 'Filtered interactions' tab gives you the result details in a table that can be searched and sorted. It also gives detailed feed-back on the characteristics of the underlying data and the SFINX analysis. The 'Distribution' tab shows you the distribution of the different SFINX scores and the effect of the 'Strictness' filter. The 'Bait data' and 'Original data' tabs allow the check-up of the input data, and the 'Network' tab allows the visualization of the filtered interactions in a network. Baits are dark blue; identified interactors are light blue.

Further information and advanced analysis

You can find further information in the 'Info' and 'About' pages.

Although the developers discourage the use of the advance version of SFINX that allows parameter optimization, the user can find this version at the 'Advanced' page. The use of parameter optimization should always be reported upon any communication of the results.

Please cite the SFINX publication upon use:

Titeca, K. et al. SFINX: straightforward filtering index for affinity purification-mass spectrometry data analysis. J Proteome Res (2015).

The SFINX algorithm and this interface were designed and created by Kevin Titeca .

He started the development of SFINX during his PhD that is guided by his promotor Jan Tavernier and co-promotor Sven Eyckerman . He is further also advised by Pieter Meysman , Kris Laukens , Kris Gevaert and Lennart Martens .

This is version 1.07 of SFINX, and is for academic use only.

SFINX is written in the R language, the web interface is partly build with Shiny, and the network visualization uses the networkD3 package for the creation of D3 JavaScript network graphs. The example data is the TIP49 data set derived from the publication by Sardiu et al.

R Core Team (2013). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. http://www.R-project.org/

Winston Chang (2015). shiny: Web Application Framework for R. R package version 0.11. http://CRAN.R-project.org/package=shiny

Christopher Gandrud, J.J. Allaire and B.W. Lewis (2014). networkD3: Tools for Creating D3 JavaScript Network Graphs from R. R package version 0.1.1. http://CRAN.R-project.org/package=networkD3

Sardiu, M.E. et al. Proc. Natl. Acad. Sci. USA 105, 1454-1459 (2008)

For questions about SFINX, you can e-mail sfinxinteractomics(at)gmail.com . Note that it can take some time before we answer.

Update details:

New in version 1.07 (09 Feb 2016): More compact possibilities for display of tables representing the original data.

New in version 1.06 (08 Feb 2016): Introduction of the loading bar for the analysis of large datasets, and enhanced user feed-back possibilities.

New in version 1.05 (07 Feb 2016): Enhancements of the interface to better handle large data analysis with many baits.

New in version 1.04 (13 Jan 2016): Better error handling and more specific feed-back possibilities.

New in version 1.03 (20 Dec 2015): Restriction of the possible datatypes for standard input.

New in version 1.02 (20 Dec 2015): Better error handling and more specific feed-back possibilities.

New in version 1.01 (14 Dec 2015): Possibility to use alternative family wise error rate corrections in the advanced version of SFINX.


Basic data (Example file)

Bait identities (Example file)
Download filtered data as tsv: Complete Output Cut-off Output

We discourage the use of this SFINX version for standard data analysis. Nevertheless, if you are a user that likes to experiment with the underlying parameters of SFINX or with bait-dependencies, then this is the place to be.

The use of parameter optimization should always be reported upon any communication of the results!
Input of example data

Check the 'Use examples' check-box in the panel at the left. All example data should load automatically.

Input of your own data

SFINX requires two files as input: a 'Basic data' file, which is a matrix with your input data, and a 'Bait identities' file, which is a list with your baits of interest. For each of the files, you can indicate whether the file has a header, and how the file is formatted: tab-separated, comma-separated or semicolon-separated.

The 'Basic data' file is a matrix with a column for each project and a row for every protein that was ever detected. The cells of the matrix are populated with the observed peptide counts for the proteins per project. (For an example: see the 'Original data' tab for the analysis of the example data.)

The baits of interest always have to be present as found proteins in your 'Basic data' file. If this is not the case, SFINX will not generate any output for these baits.

Analysis and download possibilities

All the features of the standard SFINX analysis are also present in this advanced version of SFINX. Hence, for any information about the standard features of SFINX, you can consult the short description in the 'Analysis' page or the more extensive description in the 'Info' page.

The 'background ratio' determines the ratio between the amount of bait projects and the amount of true negative controls. For example, if the ratio equals 5, then maximally 5 times more true negative controls are used. Increasing this ratio has generally very limited or no impact on the analysis, diminishing this ratio can sometimes have a bigger impact but is discouraged as it makes SFINX less strict.

If you check the 'lower cross-influence of baits' checkbox, SFINX will analyze every bait separately from the other baits. Hence, it will also avoid other bait projects as negative controls. This parametrical change is only potentially interesting, if the baits are very connected to each other.

After input of the data, you can refine the choice of negative controls. You can indicate any column of the data matrix to be used as potential negative control if that column is not already getting used as a bait project. You can also combine columns or choose the 'automatic' option in which SFINX determines the negative controls automatically.

Further information

You can find further information in the 'Info' and 'About' pages.