From 13f453718d8acdd0d7a22b0ce824e0bed01e4e2a Mon Sep 17 00:00:00 2001
From: coolneng <akasroua@gmail.com>
Date: Sat, 27 Mar 2021 09:36:59 +0100
Subject: [PATCH] Implement HVR sequence alignment

---
 src/alignment.r | 24 ++++++++++++++++++++----
 1 file changed, 20 insertions(+), 4 deletions(-)

diff --git a/src/alignment.r b/src/alignment.r
index a0a2d57..596c06a 100644
--- a/src/alignment.r
+++ b/src/alignment.r
@@ -5,7 +5,8 @@ construct_dataframe <- function(data) {
   vdj_string_set <- lapply(data, FUN = Biostrings::DNAStringSet)
   vdj_dataframe <- as.data.frame(vdj_string_set)
   vdj_dataframe$hvr_region <- paste(vdj_dataframe$v_sequence,
-    vdj_dataframe$d_sequence, vdj_dataframe$j_sequence,
+    vdj_dataframe$d_sequence,
+    vdj_dataframe$j_sequence,
     sep = ""
   )
   return(vdj_dataframe)
@@ -18,11 +19,26 @@ parse_data <- function(files) {
   vdj_dataframe <- construct_dataframe(vdj_alignment)
   return(list(sequences, vdj_dataframe))
 }
+
+align_sequence <- function(sequence, vdj_segment) {
+  return(Biostrings::pairwiseAlignment(
+    pattern = sequence,
+    subject = vdj_segment,
+    type = "global-local",
+    gapOpening = 1
+  ))
 }
 
-align_sequences <- function(sequences, vdj_segments) {
-
+perform_alignment <- function(sequences, vdj_segments) {
+  sequence_alignment <- mcmapply(sequences,
+    vdj_segments$hvr_region,
+    FUN = align_sequence,
+    mc.cores = 4
+  )
+  return(sequence_alignment)
 }
 
 input_files <- c("data/curesim_sequence.fastq", "data/vdj_alignment.csv")
-data <- parse_data(input_files)
\ No newline at end of file
+data <- parse_data(input_files)
+alignment <- perform_alignment(sequences = data[[1]], vdj_segments = data[[2]])
+print(alignment)
\ No newline at end of file