From 9c7df2262087782615bdcfbec0e15aa1a403d9e7 Mon Sep 17 00:00:00 2001
From: Gaurav Sablok <codeprog@icloud.com>
Date: Wed, 18 Sep 2024 22:29:38 +0200
Subject: [PATCH] codecreatede

coded the first draft and started the debug to fix one bug and add then the usage.
---
 README.md                        |  9 +++
 go.mod                           |  3 +
 main.go                          | 97 ++++++++++++++++++++++++++++++++
 samplefile/samplealignment.fasta |  4 ++
 4 files changed, 113 insertions(+)
 create mode 100644 README.md
 create mode 100644 go.mod
 create mode 100644 main.go
 create mode 100644 samplefile/samplealignment.fasta

diff --git a/README.md b/README.md
new file mode 100644
index 0000000..82dcead
--- /dev/null
+++ b/README.md
@@ -0,0 +1,9 @@
+go-whole-genome-estimates
+
+- a whole genome linear block estimation. 
+- block is defined as a single line for the aligned genome. 
+- no matter how many genome you aligned you can pass the alignment file.
+- the first genome will be taken as a reference. 
+- check the AlignmentGO package for defining your own block size. 
+
+Gaurav Sablok
diff --git a/go.mod b/go.mod
new file mode 100644
index 0000000..07e7592
--- /dev/null
+++ b/go.mod
@@ -0,0 +1,3 @@
+module github.com/whole-genome-alignment-estimate
+
+go 1.23.1
diff --git a/main.go b/main.go
new file mode 100644
index 0000000..1ccb019
--- /dev/null
+++ b/main.go
@@ -0,0 +1,97 @@
+package main
+
+import (
+	"bufio"
+	"flag"
+	"fmt"
+	"log"
+	"os"
+	"strconv"
+	"strings"
+)
+
+/*
+ *
+ Author Gaurav Sablok
+ Universitat Potsdam
+ Date 2024-9-19
+
+It takes a whole genome aligned files and then estimates the linear block by using the first genome as a reference genome.
+You can pass as many genome as a reference genome and it will take the first genome as a reference. This only compares the
+single line block. In the alignmentGO package, you can find the specific for defining your own block.
+
+ */
+
+ func main () {
+
+ alignment := flag.String("alignment", "path to the alignment file", "file")
+
+ flag.Parse()
+
+	type alignmentIDStore struct {
+		id string
+	}
+
+	type alignmentSeqStore struct {
+		seq string
+	}
+
+	fOpen, err := os.Open(*alignment)
+	if err != nil {
+		log.Fatal(err)
+	}
+
+	alignmentID := []alignmentIDStore{}
+	alignmentSeq := []alignmentSeqStore{}
+	sequenceSpec := []string{}
+
+	fRead := bufio.NewScanner(fOpen)
+	for fRead.Scan() {
+		line := fRead.Text()
+		if strings.HasPrefix(string(line), ">") {
+			alignmentID = append(alignmentID, alignmentIDStore{
+				id: strings.Replace(string(line), ">", "", -1),
+			})
+		}
+		if !strings.HasPrefix(string(line), ">") {
+			alignmentSeq = append(alignmentSeq, alignmentSeqStore{
+				seq: string(line),
+			})
+		}
+		if !strings.HasPrefix(string(line), ">") {
+			sequenceSpec = append(sequenceSpec, string(line))
+		}
+	}
+
+	counterA := 0
+	counterT := 0
+	counterG := 0
+	counterC := 0
+
+	for i := 0; i < len(sequenceSpec)-1; i++ {
+		for j := 0; j < len(sequenceSpec[0]); j++ {
+			if string(sequenceSpec[i][j]) == "A" && string(sequenceSpec[i+1][j]) == "T",
+			 || string(sequenceSpec[i+1][j] == "C" || string(sequenceSpec[i+1][j]) == "G" {
+					counterA++
+				}
+				if string(sequenceSpec[i][j]) == "T" && string(sequenceSpec[i+1][j]) == "C",
+					|| string(sequenceSpec[i+1][j]) == "G" || string(sequenceSpec[i+1][j]) == "A" {
+						counterT++
+					}
+					if string(sequenceSpec[i][j]) == "C" && string(sequenceSpec[i+1][j]) == "G",
+						|| string(sequenceSpec[i+1][j]) == "A" || string(sequenceSpec[i+1][j]) == "T" {
+							counterC++
+						}
+						if string(sequenceSpec[i][j]) == "G" && string(sequenceSpec[i+1][j]) == "A",
+							|| string(SequenceSpec[i+1][j]) == "T" || string(sequenceSpec[i+1][j]) == "C" {
+								counterG++
+							} else {
+								continue
+							}
+		}
+	}
+	fmt.Printf("The collinearity block of line width 1 for A as a base pattern is %d\n", counterA)
+	fmt.Printf("The collinearity block of line width 1 for T as a base pattern is %d\n", counterT)
+	fmt.Printf("The collinearity block of line width 1 for G as a base pattern is %d\n", counterG)
+	fmt.Printf("The collinearity block of line width 1 for C as a base pattern is %d\n", counterC)
+}
diff --git a/samplefile/samplealignment.fasta b/samplefile/samplealignment.fasta
new file mode 100644
index 0000000..8856d68
--- /dev/null
+++ b/samplefile/samplealignment.fasta
@@ -0,0 +1,4 @@
+>ENA|OX291461|OX291461.1
+TCAGTATC----TC--------
+>ENA|OX291509|OX291509.1
+TATC----TC--------ATAG