https://github.com/karenxzr/multinep
MultiNEP: a Multi-omics Network Enhancement framework for Prioritizing disease genes and metabolites simultanuously
https://github.com/karenxzr/multinep
multi-omics network-analysis signal-prioritization
Last synced: 10 months ago
JSON representation
MultiNEP: a Multi-omics Network Enhancement framework for Prioritizing disease genes and metabolites simultanuously
- Host: GitHub
- URL: https://github.com/karenxzr/multinep
- Owner: Karenxzr
- License: mit
- Created: 2022-03-18T22:42:02.000Z (over 4 years ago)
- Default Branch: main
- Last Pushed: 2023-06-09T17:32:15.000Z (about 3 years ago)
- Last Synced: 2025-01-30T10:35:06.230Z (over 1 year ago)
- Topics: multi-omics, network-analysis, signal-prioritization
- Language: R
- Homepage:
- Size: 8.02 MB
- Stars: 2
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# MultiNEP: a Multi-omics Network Enhancement framework for Prioritizing disease genes and metabolites simultanuously
MultiNEP is an improved analytical tool to prioritize disease-associated genes and metabolites simultanuously using multi-omics network with the ability to handle network imbalance. Multinep first reweight a general multi-omics network $S^0$ from database and a multi-omics similarity matrix $E$ based on disease profiles into $\tilde{S^0}$ and $\tilde{E}$ using weighting parameters $\lambda_g$ and $\lambda_m$. Then using reweighted $\tilde{E}$ to enhance reweighted $\tilde{S^0}$ into a disease-specific network $S_E$. At last, update initial disease-association gene and metabolite scores by diffusing on the enhanced and denoised multi-omics network $S_E$, and prioritize candidate disease-associated genes and metabolites simultanuously using updated disease-association gene and metabolite scores.
## Installation
- The R package of MultiNEP can be installed through:
`if (!requireNamespace("devtools", quietly = TRUE))`
`install.packages("devtools")`
`library("devtools")`
`install_github("Karenxzr/MultiNEP")`
## Usage
It is quite simple to run MultiNEP through a wrapper function of `nep`.
Input required:
- General network `s0`: an $n \times n$ matrix. With rownames and colnames being set as gene/metabolite names.
- Disease similarity matrix `E`: an $n \times n$ matrix. With rownames and colnames being set as gene/metabolite names. Note, all values in E should range from 0 - 1.
- Initial disease association scores `signal`: a dataframe with the first column being feature names, the second column being initial association scores. Input p-values as default association scores.
- Feature name list `feature_name_list`: a list with the first element containing all gene names and the second containing all metabolite names.
You can find sample input data within pacakge or in the `data` folder. See an application example as below:
`library(MultiNEP)`
`results = nep(s0=s0,E=E,signals=signal,feature_name_list = feature_name_list, model='multinep')`
Run `results$vec` to get prioritized candidate disease-associated multi-omics features. If you want to get re-weighted and enhanced disease-specific multi-omics network $S_E$, run `results$enhanced_mat$unprocessed` or `results$enhanced_mat$processed` with `return_mat` argument set as TRUE.
You can also change parameters such as $\lambda_g$ or $\lambda_m$. Run `?nep` to find more details.
## Reference
- Zhuoran Xu, Luigi Marchionni, Shuang Wang, MultiNEP: a multi-omics network enhancement framework for prioritizing disease genes and metabolites simultaneously, Bioinformatics, Volume 39, Issue 6, June 2023, btad333, https://doi.org/10.1093/bioinformatics/btad333