Help and documentation
WeSA, (weighted socioaffinity), is a statistical score which can rank results from protein
experiments of the bait-prey type, e.g. affinity purification, immunoprecipitation or proximity labelling.
You can enter lists of interacting proteins according to the results of your experiment.
Input your raw data before any filtering is done.
As a result you will get a re-ordered list of scores which allows one
to understand which interactions are more likely than others based on what previous experiments can add to your data.
WeSA calculations require protein pair data to
compute the scores of interactions. There are two possibilities: calculating the score using user input data or
displaying the scores as they are computed from databases. Users can provide their input in
different ways but always using gene names.
You can input a list of proteins (one protein per line) and
it will look for all interactions involving any protein on the list
and display the scores as calculated from the data in the selected database
(IntAct, BioGRID, BioPlex or any combination of the three).
If you want to re-calculate the scores while including your own experimental
results, then enter a list of protein pairs. In case you are adding results
from a single experiment, you can add a simple protein pair list (two proteins in the same line
space-separated). Alternatively, for results from multiple experiments,
each line should contain a protein pair and the experiment ID in which it was identified.
In all cases you should select a database, the data in which most closely resembles your experimental design.
IntAct is selected as default as it has more accurate records than the other two, but it is missing a lot of information
which can be found elsewhere.
Currently, it supports human proteins only.
The output is a table of protein pairs together with their calculated WeSA scores. The table is interactive
and the user is encouraged to manually explore
their results using the search and sort buttons.
The resulting table can be exported in several file formats (SVG, Excel,
PDF).
In addition to the overall WeSA score, the output table contains the observed numbers for each protein pair,
meaning, the total count of experiments in the data, in which the pair was found
(separately for the three cases: 1. spoke A - when the first protein was bait in the experiment and retrieved the other protein,
2. spoke B - when the second protein was bait and retrieved the first, 3. matrix - when the two proteins were captured
in an experiment involving another protein distinct from the two proteins of interest.
Option 1: Query the existing data
Input: Single-column list
You can input a list of proteins (one protein per line) and the webtool will output all
interactions involving those proteins and their pre-computed WeSA scores.
Sample input:
IFT20
BBS4
LZTFL1
If you do not select a dataset from the dropdown menu, the default is to display interactions
discovered and recorded within IntAct (and WeSA scores calculated from IntAct data).
| Protein A | Protein B | WeSA | SA | Count: A retrieved B | Count: B retrieved A | Count: matrix |
|---|---|---|---|---|---|---|
| BBS4 | BBS1 | 71.6 | 16.9 | 5 | 1 | 12 |
| IFT172 | IFT20 | 65.8 | 11.5 | 3 | 0 | 14 |
| BBS4 | BBS2 | 63.6 | 16 | 3 | 2 | 13 |
| LZTFL1 | BBS9 | 46.2 | 0 | 4 | 2 | 0 |
| IFT88 | IFT20 | 58.8 | 10.9 | 5 | 0 | 9 |
| ... | ||||||
Option 2: Re-calculate the scores to rank the results from your own single experiment
If you performed a single experiment (for more than one experiment refer to option 3) and are
wondering how to filter the list of your results, you can use this tool to re-calculate the WeSA
scores to rank your results. To do this, enter a list of protein pairs (two proteins in
the same line space-separated).
Input: Two-column space-separated list
The first column in every line should be the tagged bait protein and that should be separated
via space from the second protein. In this case, the second column contains all proteins which
were identified in the precipitate.
Sample input:
IFT20 IFT72
IFT20 IFT81
IFT20 IFT74
If you do not select a dataset from the dropdown menu, the default is to use interactions
discovered and recorded within IntAct (and WeSA scores calculated from IntAct data enriched with
the submitted data).
| Protein A | Protein B | WeSA | SA | Count: A retrieved B | Count: B retrieved A | Count: matrix |
|---|---|---|---|---|---|---|
| IFT20 | IFT72 | 10.6 | 10.6 | 1 | 0 | 0 |
| IFT20 | IFT74 | 8.5 | 8.5 | 1 | 0 | 0 |
| IFT20 | IFT81 | 6.9 | 6.9 | 1 | 0 | 0 |
Important note: The results only show the scores of the inputted pairs as long as evidence supporting the respective interaction exists in the chosen background dataset used for computations.
Option 3: Re-calculate the scores to rank the results from your own experiments
If you performed multiple experiments and are wondering how to filter the list of your results,
you can also use this tool to re-calculate the WeSA scores and rank your results.
To do this, enter a list of protein pairs alongside unique experiment identifiers.
Input: Three-column space-separated list
The first column in every line should be the tagged bait protein and that should be separated
via space from the second protein. In this case, the second column contains proteins which
were identified in the precipitate of the respective bait. The third entry in every line
should be the unique experiment identifier. This identifier can be anything as long as it is
unique for every separate experiment.
Sample input:
IFT20 IFT80 E1
IFT20 IFT81 E1
IFT20 IFT74 E2
IFT20 IFT80 E2
IFT80 IFT81 E3
IFT80 IFT172 E3
IFT80 IFT57 E3
IFT80 IFT46 E3
If you do not select a dataset from the dropdown menu, the default is to use interactions
discovered and recorded within IntAct (and WeSA scores calculated from IntAct data enriched with
the submitted data).
| Protein A | Protein B | WeSA | SA | Count: A retrieved B | Count: B retrieved A | Count: matrix |
|---|---|---|---|---|---|---|
| IFT80 | IFT20 | 64 | 18.8 | 2 | 1 | 14 |
| IFT20 | IFT80 | 8.1 | 8.1 | 1 | 0 | 0 |
| IFT80 | IFT172 | 8 | 8 | 1 | 0 | 0 |
| IFT80 | IFT46 | 7.9 | 7.9 | 1 | 0 | 0 |
| IFT80 | IFT57 | 7.9 | 7.9 | 1 | 0 | 0 |
| IFT20 | IFT81 | 7.6 | 7.6 | 1 | 0 | 0 |
| IFT20 | IFT74 | 7.4 | 7.4 | 1 | 0 | 0 |
| IFT80 | IFT81 | 7.3 | 7.3 | 1 | 0 | 0 |
Important note: The results only show the scores of the inputted pairs as long as evidence supporting the respective interaction exists in the chosen background dataset used for computations.
Protein A, Protein B: Two proteins which have some evidence supporting their interaction;
WeSA: WeSA score for the respective protein pair, computed from the selected dataset: IntAct, BioGRID, BioPlex or a combination. Higher WeSA scores correspond to interactions which are better supported by the evidence. The visualisation distinguishes red and green interactions: the red interactions have scores below an optimal threshold which is specific, unique and pre-computed for each dataset based on the optimal balance between true positives and false positives; the green edges correspond to WeSA scores above the optimal threshold.;
SA: Socioaffinity score for the respective protein pair;
Count: A retrieved B: Count of recorded experiments in which protein A was the tagged bait protein and protein B was retrieved;
Count: B retrieved A: (Reverse situation in comparison to previous) Count of recorded experiments in which protein B was the tagged bait protein and protein A was retrieved;
Count: matrix: Count of recorded experiments in which some protein X was the tagged bait protein and proteins A and B were retrieved simultaneously (in the same experiment).
