mappings

🗺 MAPPINGS v0.1.0

Mapping and Analysis of Phosphorylation Pathways Identified through Network/Graph Signalling

MAPPINGS is a tool for the deconvolution of complex phosphorylation-based signalling datasets. It was designed to use antibody microarray datasets (eg. Kinexus KAM antibody microarrays), though any comparative dataset (control vs treated/infected) can be analysed. The MAPPINGS tool seeks to identify pathways of consistent phosphorylation signalling changes in a comparative dataset and rapidly identify lead subnetworks for subsequent validation and exploration. The program functions through a random trails-based function which is influenced by the change in signals observed. First described by J. Adderley and F. O’Donoghue in MAPPINGS v1.0, a tool for network analysis of large phospho-signalling datasets: application to host erythrocyte response to Plasmodium infection (Adderley et al., 2021).

• Figure 1: Kinases and substrates are represented as nodes (dark nodes = substrates, light nodes = kinases) and phosphorylation events are represented as edges, which are designated with the specific phosphorylation site. Edges are represented in a colour gradient from grey to green (positive edges) and grey to red (negative edges) and a size gradient which corresponds to the percentage change from the control network trails (%CFC). The effect of the edge is represented in the arrowhead (arrow = activation, square = inhibition).

Designed and programmed with love by Dr. Jack Adderley and Finn O'Donoghue.

Installation

pip install mappings

If this doesn’t work, you may not have pip installed. - Note you will also need python installed, see dependencies.

Dependencies:

• Python - this will include pip.

• pip for a guide on how to install pip.

Usage

on Linux and MacOS

mappings ARRAY_DATA_PATH OUTPUT_PATH [OPTIONS]

or Windows

python -m mappings ARRAY_DATA_PATH OUTPUT_PATH [OPTIONS]

Note: If you drag and drop your data file.csv it will fill the PATH for you.

Examples

(Default settings) python -m mappings mappings_kinase_data.csv mappings_kinase_analysis.csv
(Custom settings)  python -m mappings mappings_kinase_data.csv mappings_kinase_analysis.csv -N 10000 --errorThreshold 0.5 --panNormaliser False 

Options

Each of the following options has a default set. Therefore, they only need to be input if different values are desired.

-N or --nwalks             (INTEGER)  Number of walks, default = 1,000,000.
--errorThreshold           (INTEGER)  Error threshold used to refine data used, default = 1.0, recommended range = 0 - 1.0, (1.0 = total error is not greater than signal, 0 = no removal of high error signals) 
--lowSignalCutOff          (INTEGER)  Removal of low intensity signals, default = 1,000, recommended range = 500 - 1,500 for Kinexus antibody microarray datasets, can be move up or down depending on the desire output network size
--panNormaliser            (BOOLEAN)  Normalises signals by available Pan-specific antibody data provided. Default = Yes (normalise)
-M or --minimumTrailLength (INTEGER)  The minimum number of edges a walk is required to pass throguh to be counted as a trail, default = 3, range = 1+, reducing this will result in more complex outputs which are less focused on pathway identification
--connection_network_path  (PATH)     Network of known phosphorylation connection network (a network is provided in data\input\NetworkComplete.csv).

Input / Output Specification

Array Data

CSV file with headers:

UniprotID – (eg. Q9Y6R4) Must be accurate as this is what is used to map the dataset into the known interaction network.

AntibodyTarget – This can be in any form or left blank

Phosphosite – This is the antibodies recognised phosphosite, in the form (Y1234 or S234 or T564, combinations of Y1234+Y1235 or S235/T537 are accepted and will be split during the analysis into the individual phosphosites. If Pan-specific antibodies are included (to enable protein level normalisation) they need to be denoted with the term "Pan"

ControlMean – Mean control signal/value (mean of technical duplicates or of biological duplicates if available). Can be performed on single none replicated signal if desired.

ControlError(%) – Mean control signal/value error as a percentage (Error range between replicates / mean signal/value * 100) – if performed on a single replicate fill this column with ‘0’

TreatedMean – Mean treated/infected signal/value (mean of technical duplicates or of biological duplicates if available). Can be performed on single none replicated signal if desired.

TreatedError(%) – Mean treated/infected signal/value error as a percentage (Error range between replicates / mean signal/value * 100) – if performed on a single replicate fill this column with ‘0’

Output Data

The output will be in CSV file format with the following default headers;

Kinase	Substrate	Phosphosite	Change(%)	Log2FoldChange	SubstrateEffect

The following additional headers can be selected;

ControlEdgeUsage	TreatedEdgeUsage

--edgeUsage  (FLAG) Add edge usage numbers to output .csv, not required for cytoscape rendering

For visualisation of the output network, we recommend using Cytoscape.

• To import the network, select file - import - network from file, select the MAPPINGS analysis output. In the pop-up window click on ‘Kinase’ and select ‘Source Node (green circle), click on Substrate and select ‘Target Node’ (orange bullseye) and ‘Phosphosite’ and select ‘Interaction type' (purple triangle).

• The network will render and using the Style tab the network can be visually customised to desired design. To use our custom style Cytoscape Style.

Connection Network

The phosphorylation network used here as a backbone for the MAPPINGS analysis is an accumulated from literature reports and has been updated to include further connections and phosphorylation effects (see Adderley et al., 2021) and original network (PhosphoAtlas). Updated version may be available under data/input/NetworkComplete.csv Please email the authors, or submit a pull request to update this file with any new data. Alternatively, additional connections can be amended into the NetworkComplete.csv if desired.

Citation

• Adderley, J. D., O’Donoghue, F., Doerig, C. & Davis, S. Network analysis of large phospho-signalling datasets: application to Plasmodium-erythrocyte interactions. bioRxiv, 2021