From 8c27b6d8e28ea80ed58c03ad8c7492e7b689ae5d Mon Sep 17 00:00:00 2001
From: mbrovkin <54407985+mbrovkin@users.noreply.github.com>
Date: Wed, 26 Mar 2025 14:52:40 -0400
Subject: [PATCH 01/16] Update grn.R

added functionality to escape special characters within gene names
---
 R/grn.R | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/R/grn.R b/R/grn.R
index b4a30ee..e14d41c 100644
--- a/R/grn.R
+++ b/R/grn.R
@@ -335,6 +335,8 @@ fit_grn_models.GRNData <- function(
             }
             peak_name <- str_replace_all(p, '-', '_')
             tf_name <- str_replace_all(peak_tfs, '-', '_')
+            #some gene names contain special characters
+            tf_name <- str_replace_all(tf_name, "([:\\(\\)])", "\\\\\\1") 
             formula_str <- paste(
                 paste(peak_name, interaction_term, tf_name, sep=' '), collapse = ' + ')
             return(list(tfs=peak_tfs, frml=formula_str))
@@ -346,6 +348,7 @@ fit_grn_models.GRNData <- function(
         }
 
         target <- str_replace_all(g, '-', '_')
+        target <- str_replace_all(target, "([:\\(\\)])", "\\\\\\1")
         model_frml <- as.formula(
             paste0(target, ' ~ ', paste0(map(frml_string, function(x) x$frml),  collapse=' + '))
         )
@@ -356,7 +359,9 @@ fit_grn_models.GRNData <- function(
         gene_x <- gene_data[gene_groups, union(g, gene_tfs), drop=FALSE]
         model_mat <- as.data.frame(cbind(gene_x, peak_x))
         if (scale) model_mat <- as.data.frame(scale(as.matrix(model_mat)))
+                                      
         colnames(model_mat) <- str_replace_all(colnames(model_mat), '-', '_')
+        colnames(model_mat) <- str_replace_all(colnames(model_mat), "([:\\(\\)])", "\\\\\\1")                        
 
         log_message('Fitting model with ', nfeats, ' variables for ', g, verbose=verbose==2)
         result <- try(fit_model(
@@ -446,15 +451,13 @@ format_coefs <- function(coefs, term=':', adjust_method='fdr'){
         select(-region_, -tf_) %>%
         mutate(
             region = str_replace_all(region, '_', '-'),
-            tf = str_replace_all(tf, '_', '-'),
-            target = str_replace_all(target, '_', '-')
+            tf = str_replace_all(tf, "\\\\([:\\(\\)])", "\\1") %>% str_replace_all('_', '-'),
+            target = str_replace_all(target, "\\\\([:\\(\\)])", "\\1") %>% str_replace_all('_', '-')
         ) %>%
         select(tf, target, region, term, everything())
     return(coefs_use)
 }
 
-
-
 #' Find TF modules in regulatory network
 #'
 #' @import tidygraph

From 5f89379864b5941e5fe5eeefd01311e0768b1a1d Mon Sep 17 00:00:00 2001
From: mbrovkin <54407985+mbrovkin@users.noreply.github.com>
Date: Wed, 26 Mar 2025 15:06:04 -0400
Subject: [PATCH 02/16] Update grn.R

fix attempt #2 for the special characters non-functionality
---
 R/grn.R | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/R/grn.R b/R/grn.R
index e14d41c..eb98497 100644
--- a/R/grn.R
+++ b/R/grn.R
@@ -336,7 +336,7 @@ fit_grn_models.GRNData <- function(
             peak_name <- str_replace_all(p, '-', '_')
             tf_name <- str_replace_all(peak_tfs, '-', '_')
             #some gene names contain special characters
-            tf_name <- str_replace_all(tf_name, "([:\\(\\)])", "\\\\\\1") 
+            tf_name <- str_replace_all(tf_name, "\\\\([:\\(\\)])", "\\1") 
             formula_str <- paste(
                 paste(peak_name, interaction_term, tf_name, sep=' '), collapse = ' + ')
             return(list(tfs=peak_tfs, frml=formula_str))
@@ -348,7 +348,7 @@ fit_grn_models.GRNData <- function(
         }
 
         target <- str_replace_all(g, '-', '_')
-        target <- str_replace_all(target, "([:\\(\\)])", "\\\\\\1")
+        target <- str_replace_all(target, "\\\\([:\\(\\)])", "\\1")
         model_frml <- as.formula(
             paste0(target, ' ~ ', paste0(map(frml_string, function(x) x$frml),  collapse=' + '))
         )
@@ -361,7 +361,7 @@ fit_grn_models.GRNData <- function(
         if (scale) model_mat <- as.data.frame(scale(as.matrix(model_mat)))
                                       
         colnames(model_mat) <- str_replace_all(colnames(model_mat), '-', '_')
-        colnames(model_mat) <- str_replace_all(colnames(model_mat), "([:\\(\\)])", "\\\\\\1")                        
+        colnames(model_mat) <- str_replace_all(colnames(model_mat), "\\\\([:\\(\\)])", "\\1")                        
 
         log_message('Fitting model with ', nfeats, ' variables for ', g, verbose=verbose==2)
         result <- try(fit_model(
@@ -451,12 +451,13 @@ format_coefs <- function(coefs, term=':', adjust_method='fdr'){
         select(-region_, -tf_) %>%
         mutate(
             region = str_replace_all(region, '_', '-'),
-            tf = str_replace_all(tf, "\\\\([:\\(\\)])", "\\1") %>% str_replace_all('_', '-'),
-            target = str_replace_all(target, "\\\\([:\\(\\)])", "\\1") %>% str_replace_all('_', '-')
+            tf = str_replace_all(tf, "([:\\(\\)])", "\\\\\\1") %>% str_replace_all('_', '-'),
+            target = str_replace_all(target, "([:\\(\\)])", "\\\\\\1") %>% str_replace_all('_', '-')
         ) %>%
         select(tf, target, region, term, everything())
     return(coefs_use)
 }
+                   
 
 #' Find TF modules in regulatory network
 #'

From 6654e416e2f99e046b938aad2e55e3f3b589d1bd Mon Sep 17 00:00:00 2001
From: mbrovkin <54407985+mbrovkin@users.noreply.github.com>
Date: Wed, 26 Mar 2025 15:25:36 -0400
Subject: [PATCH 03/16] Update grn.R

---
 R/grn.R | 14 +++++---------
 1 file changed, 5 insertions(+), 9 deletions(-)

diff --git a/R/grn.R b/R/grn.R
index eb98497..740f5d9 100644
--- a/R/grn.R
+++ b/R/grn.R
@@ -334,9 +334,7 @@ fit_grn_models.GRNData <- function(
                 return()
             }
             peak_name <- str_replace_all(p, '-', '_')
-            tf_name <- str_replace_all(peak_tfs, '-', '_')
-            #some gene names contain special characters
-            tf_name <- str_replace_all(tf_name, "\\\\([:\\(\\)])", "\\1") 
+            tf_name <- str_replace_all(peak_tfs, '-', '_') %>% str_replace_all("[\\(\\)]", "\\\\\\0")  #some gene names contain special characters
             formula_str <- paste(
                 paste(peak_name, interaction_term, tf_name, sep=' '), collapse = ' + ')
             return(list(tfs=peak_tfs, frml=formula_str))
@@ -347,8 +345,7 @@ fit_grn_models.GRNData <- function(
             return()
         }
 
-        target <- str_replace_all(g, '-', '_')
-        target <- str_replace_all(target, "\\\\([:\\(\\)])", "\\1")
+        target <- str_replace_all(g, '-', '_') %>% str_replace_all("[\\(\\)]", "\\\\\\0") 
         model_frml <- as.formula(
             paste0(target, ' ~ ', paste0(map(frml_string, function(x) x$frml),  collapse=' + '))
         )
@@ -360,8 +357,7 @@ fit_grn_models.GRNData <- function(
         model_mat <- as.data.frame(cbind(gene_x, peak_x))
         if (scale) model_mat <- as.data.frame(scale(as.matrix(model_mat)))
                                       
-        colnames(model_mat) <- str_replace_all(colnames(model_mat), '-', '_')
-        colnames(model_mat) <- str_replace_all(colnames(model_mat), "\\\\([:\\(\\)])", "\\1")                        
+        colnames(model_mat) <- str_replace_all(colnames(model_mat), '-', '_') %>% str_replace_all("[\\(\\)]", "\\\\\\0")                      
 
         log_message('Fitting model with ', nfeats, ' variables for ', g, verbose=verbose==2)
         result <- try(fit_model(
@@ -451,8 +447,8 @@ format_coefs <- function(coefs, term=':', adjust_method='fdr'){
         select(-region_, -tf_) %>%
         mutate(
             region = str_replace_all(region, '_', '-'),
-            tf = str_replace_all(tf, "([:\\(\\)])", "\\\\\\1") %>% str_replace_all('_', '-'),
-            target = str_replace_all(target, "([:\\(\\)])", "\\\\\\1") %>% str_replace_all('_', '-')
+            tf = str_replace_all(tf, '_', '-'),
+            target = str_replace_all(target, '_', '-')
         ) %>%
         select(tf, target, region, term, everything())
     return(coefs_use)

From e54889865e09e8e194879fc6348715de15dd1b0c Mon Sep 17 00:00:00 2001
From: mbrovkin <54407985+mbrovkin@users.noreply.github.com>
Date: Wed, 26 Mar 2025 15:35:18 -0400
Subject: [PATCH 04/16] Update grn.R

---
 R/grn.R | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/R/grn.R b/R/grn.R
index 740f5d9..59da8a8 100644
--- a/R/grn.R
+++ b/R/grn.R
@@ -334,7 +334,7 @@ fit_grn_models.GRNData <- function(
                 return()
             }
             peak_name <- str_replace_all(p, '-', '_')
-            tf_name <- str_replace_all(peak_tfs, '-', '_') %>% str_replace_all("[\\(\\)]", "\\\\\\0")  #some gene names contain special characters
+            tf_name <- str_replace_all(peak_tfs, '-', '_') %>% str_replace_all("[\\(\\)]", "\\\\\\\\0")   #some gene names contain special characters
             formula_str <- paste(
                 paste(peak_name, interaction_term, tf_name, sep=' '), collapse = ' + ')
             return(list(tfs=peak_tfs, frml=formula_str))
@@ -345,11 +345,11 @@ fit_grn_models.GRNData <- function(
             return()
         }
 
-        target <- str_replace_all(g, '-', '_') %>% str_replace_all("[\\(\\)]", "\\\\\\0") 
+        target <- str_replace_all(g, '-', '_') %>% str_replace_all("[\\(\\)]", "\\\\\\\\0")  
         model_frml <- as.formula(
             paste0(target, ' ~ ', paste0(map(frml_string, function(x) x$frml),  collapse=' + '))
         )
-
+        
         # Get expression data
         nfeats <- sum(map_dbl(frml_string, function(x) length(x$tfs)))
         gene_tfs <- purrr::reduce(map(frml_string, function(x) x$tfs), union)
@@ -357,8 +357,8 @@ fit_grn_models.GRNData <- function(
         model_mat <- as.data.frame(cbind(gene_x, peak_x))
         if (scale) model_mat <- as.data.frame(scale(as.matrix(model_mat)))
                                       
-        colnames(model_mat) <- str_replace_all(colnames(model_mat), '-', '_') %>% str_replace_all("[\\(\\)]", "\\\\\\0")                      
-
+        colnames(model_mat) <- str_replace_all(colnames(model_mat), '-', '_') %>% 
+                                      str_replace_all("[\\(\\)]", "\\\\\\\\0")                        
         log_message('Fitting model with ', nfeats, ' variables for ', g, verbose=verbose==2)
         result <- try(fit_model(
             model_frml,

From d466a22838dd46d05499a5428b7b374e78480e19 Mon Sep 17 00:00:00 2001
From: mbrovkin <54407985+mbrovkin@users.noreply.github.com>
Date: Wed, 26 Mar 2025 15:45:03 -0400
Subject: [PATCH 05/16] Update grn.R

---
 R/grn.R | 12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)

diff --git a/R/grn.R b/R/grn.R
index 59da8a8..b240b64 100644
--- a/R/grn.R
+++ b/R/grn.R
@@ -334,7 +334,7 @@ fit_grn_models.GRNData <- function(
                 return()
             }
             peak_name <- str_replace_all(p, '-', '_')
-            tf_name <- str_replace_all(peak_tfs, '-', '_') %>% str_replace_all("[\\(\\)]", "\\\\\\\\0")   #some gene names contain special characters
+            tf_name <- str_replace_all(peak_tfs, '-', '_') %>% str_replace_all("[\\(\\)]", "\\\\\\0")   #some gene names contain special characters
             formula_str <- paste(
                 paste(peak_name, interaction_term, tf_name, sep=' '), collapse = ' + ')
             return(list(tfs=peak_tfs, frml=formula_str))
@@ -345,7 +345,7 @@ fit_grn_models.GRNData <- function(
             return()
         }
 
-        target <- str_replace_all(g, '-', '_') %>% str_replace_all("[\\(\\)]", "\\\\\\\\0")  
+        target <- str_replace_all(g, '-', '_') %>% str_replace_all("[\\(\\)]", "\\\\\\0")  
         model_frml <- as.formula(
             paste0(target, ' ~ ', paste0(map(frml_string, function(x) x$frml),  collapse=' + '))
         )
@@ -358,7 +358,7 @@ fit_grn_models.GRNData <- function(
         if (scale) model_mat <- as.data.frame(scale(as.matrix(model_mat)))
                                       
         colnames(model_mat) <- str_replace_all(colnames(model_mat), '-', '_') %>% 
-                                      str_replace_all("[\\(\\)]", "\\\\\\\\0")                        
+                                      str_replace_all("[\\(\\)]", "\\\\\\0")                        
         log_message('Fitting model with ', nfeats, ' variables for ', g, verbose=verbose==2)
         result <- try(fit_model(
             model_frml,
@@ -447,8 +447,10 @@ format_coefs <- function(coefs, term=':', adjust_method='fdr'){
         select(-region_, -tf_) %>%
         mutate(
             region = str_replace_all(region, '_', '-'),
-            tf = str_replace_all(tf, '_', '-'),
-            target = str_replace_all(target, '_', '-')
+            tf = str_replace_all(tf, '_', '-') %>%
+                   str_replace_all("\\\\([\\(\\)])", "\\1"),
+            target = str_replace_all(target, '_', '-') %>%
+                   str_replace_all("\\\\([\\(\\)])", "\\1")
         ) %>%
         select(tf, target, region, term, everything())
     return(coefs_use)

From bcc32453ccfa2f85e844afe452b2d65ada72024c Mon Sep 17 00:00:00 2001
From: mbrovkin <54407985+mbrovkin@users.noreply.github.com>
Date: Wed, 26 Mar 2025 16:16:19 -0400
Subject: [PATCH 06/16] Update grn.R

---
 R/grn.R | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/R/grn.R b/R/grn.R
index b240b64..a6b9edb 100644
--- a/R/grn.R
+++ b/R/grn.R
@@ -345,10 +345,18 @@ fit_grn_models.GRNData <- function(
             return()
         }
 
+        
         target <- str_replace_all(g, '-', '_') %>% str_replace_all("[\\(\\)]", "\\\\\\0")  
+
+        frml_string <- map(frml_string, function(x) {
+            # Escape special characters like parentheses
+            x$frml <- str_replace_all(x$frml, '[\\(\\)]', '\\\\\\0')
+            return(x)
+            })
         model_frml <- as.formula(
             paste0(target, ' ~ ', paste0(map(frml_string, function(x) x$frml),  collapse=' + '))
         )
+    
         
         # Get expression data
         nfeats <- sum(map_dbl(frml_string, function(x) length(x$tfs)))

From 3f23b6c2caaba58af117a0e25b6373a0921f14f9 Mon Sep 17 00:00:00 2001
From: mbrovkin <54407985+mbrovkin@users.noreply.github.com>
Date: Wed, 26 Mar 2025 16:25:42 -0400
Subject: [PATCH 07/16] Update grn.R

---
 R/grn.R | 21 +++++++++------------
 1 file changed, 9 insertions(+), 12 deletions(-)

diff --git a/R/grn.R b/R/grn.R
index a6b9edb..636d1a6 100644
--- a/R/grn.R
+++ b/R/grn.R
@@ -334,7 +334,9 @@ fit_grn_models.GRNData <- function(
                 return()
             }
             peak_name <- str_replace_all(p, '-', '_')
-            tf_name <- str_replace_all(peak_tfs, '-', '_') %>% str_replace_all("[\\(\\)]", "\\\\\\0")   #some gene names contain special characters
+            tf_name <- str_replace_all(peak_tfs, '-', '_') %>%
+                       str_replace_all("[\\(\\)]", "") %>% 
+                       str_replace_all(":", "")  #some gene names contain special characters, remove
             formula_str <- paste(
                 paste(peak_name, interaction_term, tf_name, sep=' '), collapse = ' + ')
             return(list(tfs=peak_tfs, frml=formula_str))
@@ -346,13 +348,9 @@ fit_grn_models.GRNData <- function(
         }
 
         
-        target <- str_replace_all(g, '-', '_') %>% str_replace_all("[\\(\\)]", "\\\\\\0")  
+        target <- str_replace_all(g, '-', '_') %>% str_replace_all("[\\(\\)]", "") %>% 
+                       str_replace_all(":", "")  #some gene names contain special characters, remove
 
-        frml_string <- map(frml_string, function(x) {
-            # Escape special characters like parentheses
-            x$frml <- str_replace_all(x$frml, '[\\(\\)]', '\\\\\\0')
-            return(x)
-            })
         model_frml <- as.formula(
             paste0(target, ' ~ ', paste0(map(frml_string, function(x) x$frml),  collapse=' + '))
         )
@@ -366,7 +364,8 @@ fit_grn_models.GRNData <- function(
         if (scale) model_mat <- as.data.frame(scale(as.matrix(model_mat)))
                                       
         colnames(model_mat) <- str_replace_all(colnames(model_mat), '-', '_') %>% 
-                                      str_replace_all("[\\(\\)]", "\\\\\\0")                        
+                                        str_replace_all("[\\(\\)]", "") %>% 
+                                        str_replace_all(":", "")  #some gene names contain special characters, remove                       
         log_message('Fitting model with ', nfeats, ' variables for ', g, verbose=verbose==2)
         result <- try(fit_model(
             model_frml,
@@ -455,10 +454,8 @@ format_coefs <- function(coefs, term=':', adjust_method='fdr'){
         select(-region_, -tf_) %>%
         mutate(
             region = str_replace_all(region, '_', '-'),
-            tf = str_replace_all(tf, '_', '-') %>%
-                   str_replace_all("\\\\([\\(\\)])", "\\1"),
-            target = str_replace_all(target, '_', '-') %>%
-                   str_replace_all("\\\\([\\(\\)])", "\\1")
+            tf = str_replace_all(tf, '_', '-'),
+            target = str_replace_all(target, '_', '-')
         ) %>%
         select(tf, target, region, term, everything())
     return(coefs_use)

From 087b58c39004dcc61eceeb6df326cc509f118201 Mon Sep 17 00:00:00 2001
From: mbrovkin <54407985+mbrovkin@users.noreply.github.com>
Date: Wed, 26 Mar 2025 16:36:48 -0400
Subject: [PATCH 08/16] Update grn.R

---
 R/grn.R | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)

diff --git a/R/grn.R b/R/grn.R
index 636d1a6..c36f13b 100644
--- a/R/grn.R
+++ b/R/grn.R
@@ -336,7 +336,9 @@ fit_grn_models.GRNData <- function(
             peak_name <- str_replace_all(p, '-', '_')
             tf_name <- str_replace_all(peak_tfs, '-', '_') %>%
                        str_replace_all("[\\(\\)]", "") %>% 
-                       str_replace_all(":", "")  #some gene names contain special characters, remove
+                       str_replace_all(":", "") %>%
+                        str_replace_all("^([0-9])", "X\\1")
+                         #some gene names contain special characters, remove
             formula_str <- paste(
                 paste(peak_name, interaction_term, tf_name, sep=' '), collapse = ' + ')
             return(list(tfs=peak_tfs, frml=formula_str))
@@ -348,8 +350,10 @@ fit_grn_models.GRNData <- function(
         }
 
         
-        target <- str_replace_all(g, '-', '_') %>% str_replace_all("[\\(\\)]", "") %>% 
-                       str_replace_all(":", "")  #some gene names contain special characters, remove
+        target <- str_replace_all(g, '-', '_') %>% 
+                        str_replace_all("[\\(\\)]", "") %>% 
+                           str_replace_all(":", "") %>%
+                                str_replace_all("^([0-9])", "X\\1")  #some gene names contain special characters, remove
 
         model_frml <- as.formula(
             paste0(target, ' ~ ', paste0(map(frml_string, function(x) x$frml),  collapse=' + '))
@@ -365,7 +369,8 @@ fit_grn_models.GRNData <- function(
                                       
         colnames(model_mat) <- str_replace_all(colnames(model_mat), '-', '_') %>% 
                                         str_replace_all("[\\(\\)]", "") %>% 
-                                        str_replace_all(":", "")  #some gene names contain special characters, remove                       
+                                        str_replace_all(":", "") %>%
+                                        str_replace_all("^([0-9])", "X\\1") #some gene names contain special characters, remove                       
         log_message('Fitting model with ', nfeats, ' variables for ', g, verbose=verbose==2)
         result <- try(fit_model(
             model_frml,

From 913e5cdd3f13266db28d32cd9656c66a59861a7c Mon Sep 17 00:00:00 2001
From: mbrovkin <54407985+mbrovkin@users.noreply.github.com>
Date: Wed, 26 Mar 2025 16:54:19 -0400
Subject: [PATCH 09/16] Update grn.R

---
 R/grn.R | 53 +++++++++++++++++++++--------------------------------
 1 file changed, 21 insertions(+), 32 deletions(-)

diff --git a/R/grn.R b/R/grn.R
index c36f13b..c961a38 100644
--- a/R/grn.R
+++ b/R/grn.R
@@ -276,24 +276,24 @@ fit_grn_models.GRNData <- function(
     model_fits <- map_par(features, function(g){
 
         # Select peaks near gene
-        if (!g%in%rownames(peaks2gene)){
+        if (!`g`%in%rownames(peaks2gene)){
             log_message('Warning: ', g, ' not found in EnsDb', verbose=verbose==2)
             return()
         }
-        gene_peaks <- as.logical(peaks2gene[g, ])
+        gene_peaks <- as.logical(peaks2gene[`g`, ])
         if (sum(gene_peaks)==0){
-            log_message('Warning: No peaks found near ', g, verbose=verbose==2)
+            log_message('Warning: No peaks found near ', `g`, verbose=verbose==2)
             return()
         }
 
         # Select peaks correlating with target gene expression
-        g_x <- gene_data[gene_groups, g, drop=FALSE]
+        g_x <- gene_data[gene_groups, `g`, drop=FALSE]
         peak_x <- peak_data[peak_groups, gene_peaks, drop=FALSE]
         peak_g_cor <- as(sparse_cor(peak_x, g_x), 'generalMatrix')
         peak_g_cor[is.na(peak_g_cor)] <- 0
         peaks_use <- rownames(peak_g_cor)[abs(peak_g_cor[, 1]) > peak_cor]
         if (length(peaks_use)==0){
-            log_message('Warning: No correlating peaks found for ', g, verbose=verbose==2)
+            log_message('Warning: No correlating peaks found for ', `g`, verbose=verbose==2)
             return()
         }
         peak_x <- peak_x[, peaks_use, drop=FALSE]
@@ -327,53 +327,42 @@ fit_grn_models.GRNData <- function(
             rename('tf'=1, 'corr'=2)
 
         # Filter TFs and make formula string
-        frml_string <- map(names(gene_peak_tfs), function(p){
+        `frml_string` <- map(names(gene_peak_tfs), function(p){
             peak_tfs <- gene_peak_tfs[[p]]
             peak_tfs <- peak_tfs[peak_tfs%in%tfs_use]
             if (length(peak_tfs)==0){
                 return()
             }
-            peak_name <- str_replace_all(p, '-', '_')
-            tf_name <- str_replace_all(peak_tfs, '-', '_') %>%
-                       str_replace_all("[\\(\\)]", "") %>% 
-                       str_replace_all(":", "") %>%
-                        str_replace_all("^([0-9])", "X\\1")
-                         #some gene names contain special characters, remove
+            `peak_name` <- str_replace_all(p, '-', '_')
+            `tf_name` <- str_replace_all(peak_tfs, '-', '_')
             formula_str <- paste(
-                paste(peak_name, interaction_term, tf_name, sep=' '), collapse = ' + ')
-            return(list(tfs=peak_tfs, frml=formula_str))
+                paste(`peak_name`, interaction_term, `tf_name`, sep=' '), collapse = ' + ')
+            return(list(tfs=`peak_tfs`, frml=`formula_str`))
         })
-        frml_string <- frml_string[!map_lgl(frml_string, is.null)]
-        if (length(frml_string)==0){
+        `frml_string` <- `frml_string`[!map_lgl(`frml_string`, is.null)]
+        if (length(`frml_string`)==0){
             log_message('Warning: No valid peak:TF pairs found for ', g, verbose=verbose==2)
             return()
         }
 
         
-        target <- str_replace_all(g, '-', '_') %>% 
-                        str_replace_all("[\\(\\)]", "") %>% 
-                           str_replace_all(":", "") %>%
-                                str_replace_all("^([0-9])", "X\\1")  #some gene names contain special characters, remove
-
+        target <- str_replace_all(g, '-', '_')
         model_frml <- as.formula(
-            paste0(target, ' ~ ', paste0(map(frml_string, function(x) x$frml),  collapse=' + '))
+            paste0(`target`, ' ~ ', paste0(map(`frml_string`, function(x) x$`frml`),  collapse=' + '))
         )
     
         
         # Get expression data
-        nfeats <- sum(map_dbl(frml_string, function(x) length(x$tfs)))
-        gene_tfs <- purrr::reduce(map(frml_string, function(x) x$tfs), union)
+        nfeats <- sum(map_dbl(`frml_string`, function(x) length(x$`tfs`)))
+        gene_tfs <- purrr::reduce(map(`frml_string`, function(x) x$`tfs`), union)
         gene_x <- gene_data[gene_groups, union(g, gene_tfs), drop=FALSE]
         model_mat <- as.data.frame(cbind(gene_x, peak_x))
         if (scale) model_mat <- as.data.frame(scale(as.matrix(model_mat)))
                                       
-        colnames(model_mat) <- str_replace_all(colnames(model_mat), '-', '_') %>% 
-                                        str_replace_all("[\\(\\)]", "") %>% 
-                                        str_replace_all(":", "") %>%
-                                        str_replace_all("^([0-9])", "X\\1") #some gene names contain special characters, remove                       
+        colnames(model_mat) <- str_replace_all(colnames(model_mat), '-', '_')                      
         log_message('Fitting model with ', nfeats, ' variables for ', g, verbose=verbose==2)
         result <- try(fit_model(
-            model_frml,
+            `model_frml`,
             data = model_mat,
             method = method,
             ...
@@ -394,13 +383,13 @@ fit_grn_models.GRNData <- function(
         log_message('Warning: Fitting model failed for all genes.', verbose=verbose)
     }
 
-    coefs <- map_dfr(model_fits, function(x) x$coefs, .id='target')
+    coefs <- map_dfr(model_fits, function(x) x$`coefs`, .id='target')
     coefs <- format_coefs(coefs, term=interaction_term, adjust_method=adjust_method)
-    corrs <- map_dfr(model_fits, function(x) x$corr, .id='target')
+    corrs <- map_dfr(model_fits, function(x) x$`corr`, .id='target')
     if (nrow(coefs)>0){
         coefs <- suppressMessages(left_join(coefs, corrs))
     }
-    gof <- map_dfr(model_fits, function(x) x$gof, .id='target')
+    gof <- map_dfr(model_fits, function(x) x$`gof`, .id='target')
 
     params <- list()
     params[['method']] <- method

From 814c465c52c98fee7915d52379c04e3ba049bcfa Mon Sep 17 00:00:00 2001
From: mbrovkin <54407985+mbrovkin@users.noreply.github.com>
Date: Wed, 26 Mar 2025 17:10:48 -0400
Subject: [PATCH 10/16] Update grn.R

escape problematic characters in the formula accession
---
 R/grn.R | 25 +++++++++++++------------
 1 file changed, 13 insertions(+), 12 deletions(-)

diff --git a/R/grn.R b/R/grn.R
index c961a38..f53c995 100644
--- a/R/grn.R
+++ b/R/grn.R
@@ -327,20 +327,20 @@ fit_grn_models.GRNData <- function(
             rename('tf'=1, 'corr'=2)
 
         # Filter TFs and make formula string
-        `frml_string` <- map(names(gene_peak_tfs), function(p){
+        frml_string <- map(names(gene_peak_tfs), function(p){
             peak_tfs <- gene_peak_tfs[[p]]
             peak_tfs <- peak_tfs[peak_tfs%in%tfs_use]
             if (length(peak_tfs)==0){
                 return()
             }
-            `peak_name` <- str_replace_all(p, '-', '_')
-            `tf_name` <- str_replace_all(peak_tfs, '-', '_')
+            peak_name <- str_replace_all(p, '-', '_')
+            tf_name <- str_replace_all(peak_tfs, '-', '_')
             formula_str <- paste(
-                paste(`peak_name`, interaction_term, `tf_name`, sep=' '), collapse = ' + ')
-            return(list(tfs=`peak_tfs`, frml=`formula_str`))
+                paste(peak_name, interaction_term, tf_name, sep=' '), collapse = ' + ')
+            return(list(tfs=peak_tfs, frml=formula_str))
         })
-        `frml_string` <- `frml_string`[!map_lgl(`frml_string`, is.null)]
-        if (length(`frml_string`)==0){
+        frml_string <- frml_string[!map_lgl(frml_string, is.null)]
+        if (length(frml_string)==0){
             log_message('Warning: No valid peak:TF pairs found for ', g, verbose=verbose==2)
             return()
         }
@@ -348,21 +348,22 @@ fit_grn_models.GRNData <- function(
         
         target <- str_replace_all(g, '-', '_')
         model_frml <- as.formula(
-            paste0(`target`, ' ~ ', paste0(map(`frml_string`, function(x) x$`frml`),  collapse=' + '))
+            paste0(target, ' ~ ', paste0(map(frml_string, function(x) x$`frml`),  collapse=' + '))
         )
     
         
         # Get expression data
-        nfeats <- sum(map_dbl(`frml_string`, function(x) length(x$`tfs`)))
-        gene_tfs <- purrr::reduce(map(`frml_string`, function(x) x$`tfs`), union)
+        nfeats <- sum(map_dbl(frml_string, function(x) length(x$`tfs`)))
+        gene_tfs <- purrr::reduce(map(frml_string, function(x) x$`tfs`), union)
         gene_x <- gene_data[gene_groups, union(g, gene_tfs), drop=FALSE]
         model_mat <- as.data.frame(cbind(gene_x, peak_x))
         if (scale) model_mat <- as.data.frame(scale(as.matrix(model_mat)))
                                       
-        colnames(model_mat) <- str_replace_all(colnames(model_mat), '-', '_')                      
+        colnames(model_mat) <- str_replace_all(colnames(model_mat), '-', '_')
+                                      
         log_message('Fitting model with ', nfeats, ' variables for ', g, verbose=verbose==2)
         result <- try(fit_model(
-            `model_frml`,
+            model_frml,
             data = model_mat,
             method = method,
             ...

From 834587e2d3717043b3a86b4258411a6c0bb1ad1c Mon Sep 17 00:00:00 2001
From: mbrovkin <54407985+mbrovkin@users.noreply.github.com>
Date: Wed, 26 Mar 2025 18:44:59 -0400
Subject: [PATCH 11/16] Update grn.R

---
 R/grn.R | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/R/grn.R b/R/grn.R
index f53c995..13bb164 100644
--- a/R/grn.R
+++ b/R/grn.R
@@ -335,6 +335,7 @@ fit_grn_models.GRNData <- function(
             }
             peak_name <- str_replace_all(p, '-', '_')
             tf_name <- str_replace_all(peak_tfs, '-', '_')
+            tf_name <- paste0("`", tf_name, "`")
             formula_str <- paste(
                 paste(peak_name, interaction_term, tf_name, sep=' '), collapse = ' + ')
             return(list(tfs=peak_tfs, frml=formula_str))
@@ -347,10 +348,10 @@ fit_grn_models.GRNData <- function(
 
         
         target <- str_replace_all(g, '-', '_')
+        target <- paste0("`", target, "`")
         model_frml <- as.formula(
             paste0(target, ' ~ ', paste0(map(frml_string, function(x) x$`frml`),  collapse=' + '))
         )
-    
         
         # Get expression data
         nfeats <- sum(map_dbl(frml_string, function(x) length(x$`tfs`)))
@@ -363,7 +364,7 @@ fit_grn_models.GRNData <- function(
                                       
         log_message('Fitting model with ', nfeats, ' variables for ', g, verbose=verbose==2)
         result <- try(fit_model(
-            model_frml,
+            `model_frml`,
             data = model_mat,
             method = method,
             ...
@@ -449,8 +450,8 @@ format_coefs <- function(coefs, term=':', adjust_method='fdr'){
         select(-region_, -tf_) %>%
         mutate(
             region = str_replace_all(region, '_', '-'),
-            tf = str_replace_all(tf, '_', '-'),
-            target = str_replace_all(target, '_', '-')
+            tf = str_replace_all(tf, '_', '-') %>% str_replace_all(target, '`', ''),
+            target = str_replace_all(target, '_', '-') %>% str_replace_all(target, '`', '')
         ) %>%
         select(tf, target, region, term, everything())
     return(coefs_use)

From e9baaa53f1bb40ec9d9539a703ef4c7d53f66908 Mon Sep 17 00:00:00 2001
From: mbrovkin <54407985+mbrovkin@users.noreply.github.com>
Date: Wed, 26 Mar 2025 18:58:17 -0400
Subject: [PATCH 12/16] Update grn.R

---
 R/grn.R | 19 +++++++++++--------
 1 file changed, 11 insertions(+), 8 deletions(-)

diff --git a/R/grn.R b/R/grn.R
index 13bb164..8b103ec 100644
--- a/R/grn.R
+++ b/R/grn.R
@@ -276,30 +276,31 @@ fit_grn_models.GRNData <- function(
     model_fits <- map_par(features, function(g){
 
         # Select peaks near gene
-        if (!`g`%in%rownames(peaks2gene)){
+        if (!g%in%rownames(peaks2gene)){
             log_message('Warning: ', g, ' not found in EnsDb', verbose=verbose==2)
             return()
         }
-        gene_peaks <- as.logical(peaks2gene[`g`, ])
+        gene_peaks <- as.logical(peaks2gene[g, ])
         if (sum(gene_peaks)==0){
-            log_message('Warning: No peaks found near ', `g`, verbose=verbose==2)
+            log_message('Warning: No peaks found near ', g, verbose=verbose==2)
             return()
         }
-
+        log_message('Selecting peaks correlating with target gene expression...', verbose=verbose)
         # Select peaks correlating with target gene expression
-        g_x <- gene_data[gene_groups, `g`, drop=FALSE]
+        g_x <- gene_data[gene_groups, g, drop=FALSE]
         peak_x <- peak_data[peak_groups, gene_peaks, drop=FALSE]
         peak_g_cor <- as(sparse_cor(peak_x, g_x), 'generalMatrix')
         peak_g_cor[is.na(peak_g_cor)] <- 0
         peaks_use <- rownames(peak_g_cor)[abs(peak_g_cor[, 1]) > peak_cor]
         if (length(peaks_use)==0){
-            log_message('Warning: No correlating peaks found for ', `g`, verbose=verbose==2)
+            log_message('Warning: No correlating peaks found for ', g, verbose=verbose==2)
             return()
         }
         peak_x <- peak_x[, peaks_use, drop=FALSE]
         peak_motifs <- peaks2motif[gene_peaks, , drop=FALSE][peaks_use, , drop=FALSE]
 
         # Select TFs with motifs in peaks
+        log_message('Selecting TFs with motifs in peaks...', verbose=verbose)
         gene_peak_tfs <- map(rownames(peak_motifs), function(p){
             x <- as.logical(peak_motifs[p, ])
             peak_tfs <- colMaxs(motif2tf[x, , drop=FALSE])
@@ -310,6 +311,7 @@ fit_grn_models.GRNData <- function(
         names(gene_peak_tfs) <- rownames(peak_motifs)
 
         # Check correlation of peaks with target gene
+        log_message('Check correlation of peaks with target gene...', verbose=verbose)
         gene_tfs <- purrr::reduce(gene_peak_tfs, union)
         tf_x <- gene_data[gene_groups, gene_tfs, drop=FALSE]
         tf_g_cor <- as(sparse_cor(tf_x, g_x), 'generalMatrix')
@@ -354,7 +356,8 @@ fit_grn_models.GRNData <- function(
         )
         
         # Get expression data
-        nfeats <- sum(map_dbl(frml_string, function(x) length(x$`tfs`)))
+                                        
+        nfeats <- sum(map_dbl(frml_string, function(x) length(x$`tfs`)))               
         gene_tfs <- purrr::reduce(map(frml_string, function(x) x$`tfs`), union)
         gene_x <- gene_data[gene_groups, union(g, gene_tfs), drop=FALSE]
         model_mat <- as.data.frame(cbind(gene_x, peak_x))
@@ -364,7 +367,7 @@ fit_grn_models.GRNData <- function(
                                       
         log_message('Fitting model with ', nfeats, ' variables for ', g, verbose=verbose==2)
         result <- try(fit_model(
-            `model_frml`,
+            model_frml,
             data = model_mat,
             method = method,
             ...

From 5136f52e434d9aaae89c44a831556f8e881bc82a Mon Sep 17 00:00:00 2001
From: mbrovkin <54407985+mbrovkin@users.noreply.github.com>
Date: Wed, 26 Mar 2025 19:12:35 -0400
Subject: [PATCH 13/16] Update grn.R

Update handling for special characters in gene names
---
 R/grn.R | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/R/grn.R b/R/grn.R
index 8b103ec..626c216 100644
--- a/R/grn.R
+++ b/R/grn.R
@@ -453,8 +453,8 @@ format_coefs <- function(coefs, term=':', adjust_method='fdr'){
         select(-region_, -tf_) %>%
         mutate(
             region = str_replace_all(region, '_', '-'),
-            tf = str_replace_all(tf, '_', '-') %>% str_replace_all(target, '`', ''),
-            target = str_replace_all(target, '_', '-') %>% str_replace_all(target, '`', '')
+            tf = str_replace_all(tf, '_', '-') %>% str_replace_all('`', ''),
+            target = str_replace_all(target, '_', '-') %>% str_replace_all('`', '')
         ) %>%
         select(tf, target, region, term, everything())
     return(coefs_use)

From f25e6eccc97d3d3d20c680568e5134c3c18ff820 Mon Sep 17 00:00:00 2001
From: mbrovkin <54407985+mbrovkin@users.noreply.github.com>
Date: Thu, 27 Mar 2025 00:23:11 -0400
Subject: [PATCH 14/16] Update grn.R

fixed line from object@meta.data to object@data@meta.data
---
 R/grn.R | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/R/grn.R b/R/grn.R
index 626c216..f75ba4a 100644
--- a/R/grn.R
+++ b/R/grn.R
@@ -201,7 +201,7 @@ fit_grn_models.GRNData <- function(
             group_name = aggregate_rna_col,
             verbose = FALSE
         )
-        gene_groups <- object@meta.data[[aggregate_rna_col]]
+        gene_groups <- object@data@meta.data[[aggregate_rna_col]]
     }
 
     if (is.null(aggregate_peaks_col)){

From 969ac68e50a2e57ef26b53d5d9e2cfef60528a4f Mon Sep 17 00:00:00 2001
From: mbrovkin <54407985+mbrovkin@users.noreply.github.com>
Date: Thu, 27 Mar 2025 11:38:41 -0400
Subject: [PATCH 15/16] Update grn.R

backtick peaks
---
 R/grn.R | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/R/grn.R b/R/grn.R
index f75ba4a..cf37fe4 100644
--- a/R/grn.R
+++ b/R/grn.R
@@ -336,6 +336,7 @@ fit_grn_models.GRNData <- function(
                 return()
             }
             peak_name <- str_replace_all(p, '-', '_')
+            peak_name <- paste0("`", peak_name, "`")
             tf_name <- str_replace_all(peak_tfs, '-', '_')
             tf_name <- paste0("`", tf_name, "`")
             formula_str <- paste(
@@ -452,7 +453,7 @@ format_coefs <- function(coefs, term=':', adjust_method='fdr'){
         ) %>%
         select(-region_, -tf_) %>%
         mutate(
-            region = str_replace_all(region, '_', '-'),
+            region = str_replace_all(region, '_', '-') %>% str_replace_all('`', ''),
             tf = str_replace_all(tf, '_', '-') %>% str_replace_all('`', ''),
             target = str_replace_all(target, '_', '-') %>% str_replace_all('`', '')
         ) %>%

From 36f72fa939a6b291f86eeac4b5a6c892dadf57ae Mon Sep 17 00:00:00 2001
From: mbrovkin <54407985+mbrovkin@users.noreply.github.com>
Date: Fri, 28 Mar 2025 15:01:00 -0400
Subject: [PATCH 16/16] Update grn.R

---
 R/grn.R | 1 +
 1 file changed, 1 insertion(+)

diff --git a/R/grn.R b/R/grn.R
index cf37fe4..cf3c891 100644
--- a/R/grn.R
+++ b/R/grn.R
@@ -453,6 +453,7 @@ format_coefs <- function(coefs, term=':', adjust_method='fdr'){
         ) %>%
         select(-region_, -tf_) %>%
         mutate(
+            term = str_replace_all(term, '_', '-') %>% str_replace_all('`', ''),
             region = str_replace_all(region, '_', '-') %>% str_replace_all('`', ''),
             tf = str_replace_all(tf, '_', '-') %>% str_replace_all('`', ''),
             target = str_replace_all(target, '_', '-') %>% str_replace_all('`', '')