elisabettasciacca
commented
1 year ago Hi! Thanks for your interest in the package.
1 - As you say, the analysis of all possible gene-gene combinations (of the order of 10^9) would be confounded by large numbers of false positive gene pairs biologically unrelated. Here I use a curated interaction network to address this issue and reduce the number of possible gene-gene interactions to those which are functionally relevant. This network, called meta-pathway, contains only 10537 molecular interactions (i.e. gene-gene pairs) obtained from four pathway repositories known in literature (KEGG (Kanehisa & Goto, 2000), mirTARbase (Da Hsu et al., 2011), miRecords (Xiao et al., 2009), and transmiR (Tong et al., 2019)). For more info about the meta-pathway please have a look at Alaimo et al., 2016.
2 - To use the package you need a matrix of normalized read counts where genes are organized in rows and samples in columns. Each matrix entry must represent the number of sequencing reads mapped to a gene in a sample. Data must be scaled to account for sequencing depth and heteroscedasticity (i.e. a VST normalisation would be fine). You will then need a second a second design matrix just to map samples with the subgroups you intend to use for differential comparison. For example, if you want to compare treated samples with untreated ones you will need to provide a matrix where each sample ID (indicated in the columns of the expression matrix) is classified as treated or untreated. For a complete example, please have a look at the 'Case study' paragraph of the following paper: Sciacca, Elisabetta, et al. "DEGGs: an R package with shiny app for the identification of differentially expressed gene–gene interactions in high-throughput sequencing data." Bioinformatics 39.4 (2023): btad192.
Or, install the package:
devtools::install_github("elisabettasciacca/DEGGs", build_vignettes = TRUE)
and run the vignette.
Hope that helped, let me know if you have further questions! Elisabetta
handesome
hi! Thank you for developing the DEGGs package!! how could you resolve the problem of false positive of gene-gene pairs in the R package? And how can I use my own expression data to find the true gene-gene pairs ? Thanks!!!