About
With more and more data being collected, modern network representations exploit the complementary nature of different data sources as well as similarities across patients. We here introduce the Variation of information fused Layers of Networks algorithm (ViLoN) – online. ViLoN is a novel network-based approach for the integration of multiple molecular profiles. As a key innovation, it directly incorporates prior functional knowledge (KEGG, GO). In the constructed network of patients, patients are represented by networks of pathways, comprising genes that are linked by common functions and joint regulation in the disease. ViLoN was validated for patient stratification on multiple data type combinations (gene expression, methylation, copy number), showing substantial improvements and consistently competitive performance for all (Kańduła et al., NAR 2022, gkac988). Notably, the incorporation of prior functional knowledge was critical for good results in the smaller cohorts (rectum adenocarcinoma: 90, esophageal carcinoma: 180), where alternative methods failed.
vilon.online is maintained by Alexander Aldoshin. Maciej Kańduła made key contributions to the conception and design of the web-tool and is responsible for the analyses algorithm.
This website was tested on Firefox 106.0.1, Safari 16.1, and Chrome 94.0.
MAQC Consortium 2011–2014 (read more)
FDA SEQC, Nature Biotechnology (read more)
Bioinformatics (read more)
FEBS Journal (read more)
Gerontology (read more)
ViLoN Online
Patient network clustering
File formats and Examples
Example input files and parameters
Input: ZIP archives of CSV files of normalized molecular profiles, optionally multiple profile types (gene-expression, copy-number variation, ...), plus optionally a file with survival times. Formats are documented here and through the above example input files. Specifically, row-names hold gene-names according to the HGNC (e.g., A2BP1, ANKRD23, NPDC1,...). Patient identifiers are given in the column names. The survival file lists the days to last follow-up and a status 1 for censored (=alive) or 2 for deceased (or relapsed).
Output: A folder of results for each k, including a simple spreadsheet table with patients stratified into groups. For any provided survival data we show Cox regression result tables and a metric of clinical stratification relevance for each patient cluster grouping, comprising a score of the effective number of affected patients. Optionally, comparative graphs of these results are compiled for the range of k and selected alternative methods (SNFtool, LRAcluster), and Kaplan-Meier curves can be generated.