WSSmod provides tools to construct Weighted Sum Score (WSS) module-level summaries from proteomic data. Proteins are grouped into STRING network modules, sign-flipped to align with the direction of core biomarker association, and combined into a single module score per sample using a weighted average based on network connectivity (alpha centrality) and biomarker association strength.
This package accompanies the Longo Lab preprint, Butler et al. (2026), Alzheimer’s disease blood biomarkers reveal proteomic modules of disease progression.
You can install the development version of WSSmod from GitHub with:
# install.packages("pak")
pak::pak("Longo-Lab/WSSmod")Using a custom module/weighting table:
library(WSSmod)
expr_matrix <- matrix(
c(1, 2, 3, 4, 5, 6, 7, 8),
nrow = 2,
dimnames = list(c("sample1", "sample2"), c("geneA", "geneB", "geneC", "geneD"))
)
result <- data.table::data.table(
module = c("mod1", "mod1", "mod1", "mod2"),
symbol = c("geneA", "geneB", "geneC", "geneD"),
mean_beta = c(1, -1, 1, 1),
mean_alpha_scaled = c(1, 1, 2, 1)
)
WSSmod::calculate_WSS(expr_matrix, result, min_module_size = 1)
#> $scores
#> mod1 mod2
#> sample1 2.0 7
#> sample2 2.5 8
#>
#> $summary_stats
#> module Min Q1 Median Mean Q3 Max
#> <char> <num> <num> <num> <num> <num> <num>
#> 1: mod1 2 2.125 2.25 2.25 2.375 2.5
#> 2: mod2 7 7.250 7.50 7.50 7.750 8.0
#>
#> $module_sd
#> mod1 mod2
#> 0.3535534 0.7071068
#>
#> $module_proteins
#> $module_proteins$mod1
#> [1] "geneC" "geneA" "geneB"
#>
#> $module_proteins$mod2
#> [1] "geneD"WSSmod also ships prebuilt module/weighting reference sets so you
don’t have to build your own module table from scratch.
calculate_WSS() uses
prebuilt = "core_AD_plasma_biomarkers" by default when
result = NULL:
list_prebuilt_wss()
# expr_matrix here has samples as rows and protein/gene symbols as columns,
# matching the symbols used by the prebuilt reference set
WSSmod::calculate_WSS(expr_matrix)
# Human-readable module term labels for the same reference set
load_prebuilt_terms("core_AD_plasma_biomarkers")predict_WSS() predicts module scores for a prebuilt
reference set from a small set of covariates, using a fitted
joinet model, without needing the full proteomic panel. For
"core_AD_plasma_biomarkers" this predicts all 75 module
scores from Age, Gender, and the 8 core plasma biomarkers used in the
original analysis. This requires the glmnet and
joinet packages.
The 8 biomarker columns must already be rank-based inverse-normal
transformed before calling predict_WSS().
normalize_wss_biomarkers() builds these columns (keeping
the same column names) from raw biomarker values using one of two
methods:
method = "project" (default) — project each raw value
onto a bundled reference distribution via
project_rank_norm(). Works for a single new patient or a
small/differently-distributed cohort, since each value is scored
independently against the fixed reference.method = "self" — rank-normalize
(RNOmni::RankNorm()) within your own batch, if you have a
sizeable cohort of your own with a distribution you’re comfortable
normalizing against directly instead of the bundled reference. Requires
multiple samples, since a rank transform needs multiple values to rank
against.# raw_biomarkers: one row per sample, the 8 raw (un-normalized) biomarker columns.
# Works for a single patient (method = "project" is the default):
raw_biomarkers <- data.frame(
PlasmaPTau181 = 1.5, PlasmaAB142P = 25, PlasmaAB140P = 300, PlasmaABRatio = 0.09,
PlasmapTau217 = 0.3, PlasmapTau217_AB42Ratio = 0.012, PlasmaGFAP = 60, PlasmaNfL = 22
)
normalized <- normalize_wss_biomarkers(raw_biomarkers)
newdata <- cbind(Age = 72, Gender = 1, normalized) # Gender: 1 = male, 0 = not male
pred <- predict_WSS(newdata)
pred$metaThe WSSmod code is released under the MIT license. The bundled reference files are derived from the data and results of the preprint above, which is licensed under CC BY-NC 4.0 International; please observe its terms, including non-commercial use, when reusing them.