From 3209641768fec86b3a0e338880cfdc4b7adfece5 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 | 20 ++++++++++++++++++--
 1 file changed, 18 insertions(+), 2 deletions(-)

diff --git a/src/alignment.r b/src/alignment.r
index a0a2d57..235406c 100644
--- a/src/alignment.r
+++ b/src/alignment.r
@@ -18,11 +18,27 @@ 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