From 11b5e7a68a85768c53dd1b68e1e22956b4cd70ad Mon Sep 17 00:00:00 2001
From: Gene233 <chen_jj97@163.com>
Date: Tue, 12 Nov 2024 17:35:01 +1100
Subject: [PATCH] update top_markers function, allow batch correction for glm
 estimation.

---
 DESCRIPTION            |  2 +-
 R/top_markers.R        | 15 +++++++++++++--
 man/smartid_Package.Rd |  9 +++++++++
 man/top_markers_glm.Rd |  3 +++
 4 files changed, 26 insertions(+), 3 deletions(-)

diff --git a/DESCRIPTION b/DESCRIPTION
index 87af15e..fbeabc5 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -14,7 +14,7 @@ biocViews: Software, GeneExpression, Transcriptomics
 License: MIT + file LICENSE
 Encoding: UTF-8
 Roxygen: list(markdown = TRUE)
-RoxygenNote: 7.3.1
+RoxygenNote: 7.3.2
 Collate: 
     'AllClasses.R'
     'tf_idf_iae_wrappers.R'
diff --git a/R/top_markers.R b/R/top_markers.R
index 0a506ca..cc4cb1f 100644
--- a/R/top_markers.R
+++ b/R/top_markers.R
@@ -106,6 +106,7 @@ top_markers_abs <- function(data, label, n = 10,
 #' @param data matrix, features in row and samples in column
 #' @param n integer, number of returned top genes for each group
 #' @param family family for glm, details in [stats::glm()]
+#' @param batch a vector of batch labels, default NULL
 #' @param scale logical, if to scale data by row
 #' @param use.mgm logical, if to scale data using [scale_mgm()]
 #' @param softmax logical, if to apply softmax transformation on output
@@ -119,6 +120,7 @@ top_markers_abs <- function(data, label, n = 10,
 #' top_markers_glm(data, label = rep(c("A", "B"), 5))
 top_markers_glm <- function(data, label, n = 10,
                             family = gaussian(), # score are continuous non-negative, can use gamma or inverse.gaussian, if continuous and unbounded use gaussian, if discrete use poisson, if binary or proportions between [0,1] or binary freq counts use binomial
+                            batch = NULL,
                             scale = TRUE, use.mgm = TRUE,
                             pooled.sd = FALSE,
                             # log = TRUE,
@@ -144,8 +146,17 @@ top_markers_glm <- function(data, label, n = 10,
   #   data <- log(data + 1e-8)
   # }
 
-  ## estimate betas based on given group
-  betas <- apply(data, 1, \(s) glm(s ~ 0 + label, family = family)$coef)
+  ## estimate betas based on given group and/or batch label
+  if(is.null(batch)) {
+    ## model with group label only
+    betas <- apply(data, 1, \(s) glm(s ~ 0 + label, family = family)$coef)
+  } else {
+    ## model with both group and batch label
+    betas <- apply(data, 1, \(s) glm(s ~ 0 + label + batch, family = family)$coef)
+    ## only extract betas of group label
+    betas <- betas[grep("^label", rownames(betas)), , drop = FALSE]
+  }
+
   rownames(betas) <- gsub("label", "", rownames(betas))
 
   # ## compute logFC (1 vs all mean) for each group
diff --git a/man/smartid_Package.Rd b/man/smartid_Package.Rd
index 402a787..07bbd1c 100644
--- a/man/smartid_Package.Rd
+++ b/man/smartid_Package.Rd
@@ -2,6 +2,7 @@
 % Please edit documentation in R/smartid-package.R
 \docType{package}
 \name{smartid_Package}
+\alias{smartid-package}
 \alias{smartid_Package}
 \alias{smartid}
 \title{Scoring and Marker Selection method based on modified TF-IDF}
@@ -14,6 +15,14 @@ especially for rare population. The package is developed for generating
 specific lists of signature genes based on TF-IDF modified methods. It
 can also be used as a new gene-set scoring method or data transformation
 method. Multiple visualization functions are implemented in this package.
+}
+\seealso{
+Useful links:
+\itemize{
+  \item \url{https://davislaboratory.github.io/smartid}
+  \item Report bugs at \url{https://github.com/DavisLaboratory/smartid/issues}
+}
+
 }
 \author{
 Jinjin Chen \email{chen.j@wehi.edu.au}
diff --git a/man/top_markers_glm.Rd b/man/top_markers_glm.Rd
index 43fc9d9..fe9dde8 100644
--- a/man/top_markers_glm.Rd
+++ b/man/top_markers_glm.Rd
@@ -9,6 +9,7 @@ top_markers_glm(
   label,
   n = 10,
   family = gaussian(),
+  batch = NULL,
   scale = TRUE,
   use.mgm = TRUE,
   pooled.sd = FALSE,
@@ -25,6 +26,8 @@ top_markers_glm(
 
 \item{family}{family for glm, details in \code{\link[stats:glm]{stats::glm()}}}
 
+\item{batch}{a vector of batch labels, default NULL}
+
 \item{scale}{logical, if to scale data by row}
 
 \item{use.mgm}{logical, if to scale data using \code{\link[=scale_mgm]{scale_mgm()}}}