Table 2 Qualitative comparison of SignalingProfiler 2.0 and existing methods

From: SignalingProfiler 2.0 a network-based approach to bridge multi-omics data to phenotypic hallmarks

Omics data properties

SignalingProfiler 2.0

COSMOS

CausalPath

TPS

CARNIVAL

HiPathia

CausalR

NicheNet

KPNN

Omics layers

Transcriptomics

X

X

X

 

X

X

X

X

X

Proteomics

X

 

X

      

Phosphoproteomics

X

X

X

X

     

Metabolomics

 

X

       

Biological resolution

Bulk/Pseudo-bulk

X

X

X

X

X

X

X

X

 

Single-cell

     

X

 

X

X

Omics data used as

Measurements (observations)

   

X

 

X

  

X

Statistical scores (contrast/correlation)

X

X

X

X

X

 

X

X

 

Additional input

Does not use additional inputs

 

X

   

X

X

 

X

Accepts additional inputs

X

 

X

 

X

  

X

 

Requires additional inputs

   

X

     

PKN properties

Sign

Directed

   

X

   

X

X

Activations/Inhibitions

X

X

X

X

X

X

X

  

Size

Pathways

     

X

   

Large networks

X

X

X

X

X

 

X

X

X

Biological content

Protein signaling interactions

X

X

X

X

X

  

X

X

Gene regulatory interactions

X

X

X

  

X

 

X

X

Method properties

Omics to PKN

Direct mapping to nodes

  

X

X

 

X

X

X

X

Indirect mapping to nodes

X

X

  

X

    

Not‐measured nodes

Estimates unmeasured nodes state

X

X

 

X

X

X

X

 

X

Includes unmeasured nodes in output

X

X

 

X

X

X

X

X

X

Algorithm type

Edge filtering and shortest path

X

 

X

   

X

X

 

Recursive signal propagation and heat diffusion

     

X

   

Integer linear programming

X

X

 

X

X

    

Neural networks

        

X

Final network properties

 

Includes user-defined perturbed nodes

X

   

X

X

   
 

Includes phenotypes

X

    

X

   
 

Includes meta-information on nodes

X

X

  

X

  

X

 
 

Includes meta-information on edges regarding phosphorylation events

X

 

X

X