search

glasgowcompbio / PALS

Ranking metabolite (and other omics sets) by their activity levels via SVD
https://pals.glasgowcompbio.org
MIT License
2 stars 1 forks source link

Evaluation #10

Closed joewandy closed 5 years ago

joewandy commented 5 years ago

Ronan and I discussed some ideas on how to evaluate our method for the manuscript:

  1. Run the same Colorectal cancer dataset from [1] using both the KEGG and Reactome databases in PALS. Report how we rank the same three KEGG pathways from Table 3 and the one Reactome pathway from Table 4 in [1]. Optional pre-requisite: an up-to-date KEGG database (issue #9).

  2. Implements recall and discrimination performance metrics as described in [2]. Choose any one dataset that we have (HAT?) and perform an experiment comparing PALS vs standard ORA using recall and discrimination. Pre-requisite: an implementation of ORA (issue #8).

References:

[1] Marco-Ramell, Anna, et al. "Evaluation and comparison of bioinformatic tools for the enrichment analysis of metabolomics data." BMC bioinformatics 19.1 (2018): 1. https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-017-2006-0

[2] Yu, Chenggang, et al. "A strategy for evaluating pathway analysis methods." BMC bioinformatics 18.1 (2017): 453. https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-017-1866-7

joewandy commented 5 years ago

We decided not to do 1. Instead we will do 2.