obitools/obitools4
2
0
Fork 0
mirror of https://github.com/metabarcoding/obitools4.git synced 2025-12-10 01:30:27 +00:00
Code Issues Packages Projects Releases Wiki Activity

work on obiclean chimera detection

Browse Source
This commit is contained in:
Eric Coissac
2025-10-20 16:35:19 +02:00
parent 29bf4ce871
commit d17a9520b9
6 changed files with 160 additions and 92 deletions
Download Patch File Download Diff File Expand all files Collapse all files

31
pkg/obialign/fastlcsegf.go
View File

@@ -4,33 +4,6 @@ import (
"git.metabarcoding.org/obitools/obitools4/obitools4/pkg/obiseq"
)
var _iupac = [26]byte{
// a b c d e f
1, 14, 2, 13, 0, 0,
// g h i j k l
4, 11, 0, 0, 12, 0,
// m n o p q r
3, 15, 0, 0, 0, 5,
// s t u v w x
6, 8, 8, 13, 9, 0,
// y z
10, 0,
}
func _samenuc(a, b byte) bool {
if (a >= 'A') && (a <= 'Z') {
a |= 32
}
if (b >= 'A') && (b <= 'Z') {
b |= 32
}
if (a >= 'a') && (a <= 'z') && (b >= 'a') && (b <= 'z') {
return (_iupac[a-'a'] & _iupac[b-'a']) > 0
}
return a == b
}
// FastLCSEGFScoreByte calculates the score of the Longest Common Subsequence (LCS) between two byte slices.
//
// The score is calculated using the following scoring matrix:
@@ -165,7 +138,7 @@ func FastLCSEGFScoreByte(bA, bB []byte, maxError int, endgapfree bool, buffer *[
default:
// We are in the middle of the matrix
Sdiag = _incpath(previous[x])
if _samenuc(bA[j-1], bB[i-1]) {
if obiseq.SameIUPACNuc(bA[j-1], bB[i-1]) {
Sdiag = _incscore(Sdiag)
}
@@ -265,7 +238,7 @@ func FastLCSEGFScoreByte(bA, bB []byte, maxError int, endgapfree bool, buffer *[
Sleft = _notavail
default:
Sdiag = _incpath(previous[x])
if _samenuc(bA[j-1], bB[i-1]) {
if obiseq.SameIUPACNuc(bA[j-1], bB[i-1]) {
Sdiag = _incscore(Sdiag)
}

9
pkg/obialign/locatepattern.go
View File

@@ -1,6 +1,9 @@
package obialign
import log "github.com/sirupsen/logrus"
import (
"git.metabarcoding.org/obitools/obitools4/obitools4/pkg/obiseq"
log "github.com/sirupsen/logrus"
)
// buffIndex converts a pair of coordinates (i, j) into a linear index in a matrix
// of size width x width. The coordinates are (-1)-indexed, and the linear index
@@ -69,7 +72,7 @@ func LocatePattern(id string, pattern, sequence []byte) (int, int, int) {
// Mismatch score = -1
// Match score = 0
match := -1
if _samenuc(pattern[j], sequence[i]) {
if obiseq.SameIUPACNuc(pattern[j], sequence[i]) {
match = 0
}
@@ -103,7 +106,7 @@ func LocatePattern(id string, pattern, sequence []byte) (int, int, int) {
// Mismatch score = -1
// Match score = 0
match := -1
if _samenuc(pattern[jmax], sequence[i]) {
if obiseq.SameIUPACNuc(pattern[jmax], sequence[i]) {
match = 0
}