|
| 1 | +/* Copyright (C) 2019 Philipp Benner |
| 2 | + * |
| 3 | + * This program is free software: you can redistribute it and/or modify |
| 4 | + * it under the terms of the GNU General Public License as published by |
| 5 | + * the Free Software Foundation, either version 3 of the License, or |
| 6 | + * (at your option) any later version. |
| 7 | + * |
| 8 | + * This program is distributed in the hope that it will be useful, |
| 9 | + * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 10 | + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 11 | + * GNU General Public License for more details. |
| 12 | + * |
| 13 | + * You should have received a copy of the GNU General Public License |
| 14 | + * along with this program. If not, see <http://www.gnu.org/licenses/>. |
| 15 | + */ |
| 16 | + |
| 17 | +package gonetics |
| 18 | + |
| 19 | +/* -------------------------------------------------------------------------- */ |
| 20 | + |
| 21 | +//import "fmt" |
| 22 | +import "sort" |
| 23 | +import "strings" |
| 24 | + |
| 25 | +/* -------------------------------------------------------------------------- */ |
| 26 | + |
| 27 | +type KmerCatalogue struct { |
| 28 | + KmerEquivalenceRelation |
| 29 | + names []map[int]string // k-mer names (equivalent k-mers separated by pipe) |
| 30 | + idmap []map[string]int // unique k-mer IDs |
| 31 | +} |
| 32 | + |
| 33 | +/* -------------------------------------------------------------------------- */ |
| 34 | + |
| 35 | +func NewKmerCatalogue(n, m int, comp, rev, rc bool, maxAmbiguous []int, al ComplementableAlphabet) (*KmerCatalogue, error) { |
| 36 | + r := KmerCatalogue{} |
| 37 | + if f, err := NewKmerEquivalenceRelation(n, m, comp, rev, rc, maxAmbiguous, al); err != nil { |
| 38 | + return nil, err |
| 39 | + } else { |
| 40 | + r.KmerEquivalenceRelation = f |
| 41 | + } |
| 42 | + idmap := make([]map[string]int, m-n+1) |
| 43 | + names := make([]map[int]string, m-n+1) |
| 44 | + for k := n; k <= m; k++ { |
| 45 | + idmap[k-n] = make(map[string]int) |
| 46 | + names[k-n] = make(map[int]string) |
| 47 | + } |
| 48 | + r.names = names |
| 49 | + r.idmap = idmap |
| 50 | + return &r, nil |
| 51 | +} |
| 52 | + |
| 53 | +/* -------------------------------------------------------------------------- */ |
| 54 | + |
| 55 | +func (obj *KmerCatalogue) GetId(kmer string) int { |
| 56 | + k := len(kmer) |
| 57 | + if k < obj.n || k > obj.m { |
| 58 | + panic("k-mer has invalid length") |
| 59 | + } |
| 60 | + if i, ok := obj.idmap[k-obj.n][kmer]; ok { |
| 61 | + return i |
| 62 | + } else { |
| 63 | + r := obj.EquivalenceClass(kmer) |
| 64 | + for _, kmer := range strings.Split(r.Name, "|") { |
| 65 | + obj.idmap[k-obj.n][kmer] = r.I |
| 66 | + } |
| 67 | + obj.names[k-obj.n][r.I] = r.Name |
| 68 | + return r.I |
| 69 | + } |
| 70 | +} |
| 71 | + |
| 72 | +func (obj *KmerCatalogue) GetName(kmer string) string { |
| 73 | + k := len(kmer) |
| 74 | + if k < obj.n || k > obj.m { |
| 75 | + panic("k-mer has invalid length") |
| 76 | + } |
| 77 | + if i, ok := obj.idmap[k-obj.n][kmer]; ok { |
| 78 | + return obj.names[k-obj.n][i] |
| 79 | + } else { |
| 80 | + r := obj.EquivalenceClass(kmer) |
| 81 | + for _, kmer := range strings.Split(r.Name, "|") { |
| 82 | + obj.idmap[k-obj.n][kmer] = r.I |
| 83 | + } |
| 84 | + obj.names[k-obj.n][r.I] = r.Name |
| 85 | + return r.Name |
| 86 | + } |
| 87 | +} |
| 88 | + |
| 89 | +func (obj *KmerCatalogue) IdToName(k, id int) string { |
| 90 | + if name, ok := obj.names[k-obj.n][id]; ok { |
| 91 | + return name |
| 92 | + } |
| 93 | + panic("k-mer name not found") |
| 94 | +} |
| 95 | + |
| 96 | +func (obj *KmerCatalogue) GetNames(kmer string) []string { |
| 97 | + return strings.Split(obj.GetName(kmer), "|") |
| 98 | +} |
| 99 | + |
| 100 | +func (obj *KmerCatalogue) ObservedKmers() int { |
| 101 | + r := 0 |
| 102 | + for i := 0; i < len(obj.names); i++ { |
| 103 | + r += len(obj.names[i]) |
| 104 | + } |
| 105 | + return r |
| 106 | +} |
| 107 | + |
| 108 | +/* -------------------------------------------------------------------------- */ |
| 109 | + |
| 110 | +func (obj *KmerCatalogue) scanSubKmers_(kmer []byte, k int) []int { |
| 111 | + idMap := make(map[int]struct{}) |
| 112 | + // loop over sequence |
| 113 | + for i := 0; i < len(kmer); i++ { |
| 114 | + if i+k-1 >= len(kmer) { |
| 115 | + break |
| 116 | + } |
| 117 | + it := NewKmerInstantiationIterator(obj.al, string(kmer[i:i+k]), true) |
| 118 | + for ; it.Ok(); it.Next() { |
| 119 | + if id, ok := obj.idmap[k-obj.n][it.Get()]; ok { |
| 120 | + idMap[id] = struct{}{} |
| 121 | + } |
| 122 | + } |
| 123 | + } |
| 124 | + ids := []int{} |
| 125 | + for id, _ := range idMap { |
| 126 | + ids = append(ids, id) |
| 127 | + } |
| 128 | + sort.Ints(ids) |
| 129 | + return ids |
| 130 | +} |
| 131 | + |
| 132 | +func (obj *KmerCatalogue) scanSubKmers(kmer []byte) []string { |
| 133 | + names := []string{} |
| 134 | + for k := obj.n; k <= obj.m; k++ { |
| 135 | + ids := obj.scanSubKmers_(kmer, k) |
| 136 | + for _, id := range ids { |
| 137 | + names = append(names , obj.IdToName(k, id)) |
| 138 | + } |
| 139 | + } |
| 140 | + return names |
| 141 | +} |
| 142 | + |
| 143 | +/* -------------------------------------------------------------------------- */ |
| 144 | + |
| 145 | +func (obj *KmerCatalogue) relatedKmers(s []byte, m, k int) []int { |
| 146 | + idMap := make(map[int]struct{}) |
| 147 | + // loop over positions where the k-mer can be fixed |
| 148 | + for j := 0; j <= k - len(s); j++ { |
| 149 | + for it := NewKmerCylinderIterator(k, obj.ma[k-obj.n] - m, obj.al, j, string(s)); it.Ok(); it.Next() { |
| 150 | + if id, ok := obj.idmap[k-obj.n][it.Get()]; ok { |
| 151 | + idMap[id] = struct{}{} |
| 152 | + } |
| 153 | + } |
| 154 | + } |
| 155 | + ids := []int{} |
| 156 | + for id, _ := range idMap { |
| 157 | + ids = append(ids, id) |
| 158 | + } |
| 159 | + sort.Ints(ids) |
| 160 | + return ids |
| 161 | +} |
| 162 | + |
| 163 | +func (obj *KmerCatalogue) RelatedKmers(kmer string) []string { |
| 164 | + s := []byte(obj.GetNames(kmer)[0]) |
| 165 | + m := obj.countAmbiguous(kmer) |
| 166 | + // scan k-mer for sub-k-mers |
| 167 | + names := obj.scanSubKmers(s) |
| 168 | + // loop over k-mer sizes |
| 169 | + for k := len(s)+1; k <= obj.m; k++ { |
| 170 | + ids := obj.relatedKmers(s, m, k) |
| 171 | + for _, id := range ids { |
| 172 | + names = append(names , obj.IdToName(k, id)) |
| 173 | + } |
| 174 | + } |
| 175 | + return names |
| 176 | +} |
| 177 | + |
| 178 | +func (obj *KmerCatalogue) countAmbiguous(kmer string) int { |
| 179 | + m := 0 |
| 180 | + s := []byte(kmer) |
| 181 | + for i := 0; i < len(s); i++ { |
| 182 | + if ok, err := obj.al.IsAmbiguous(s[i]); err != nil { |
| 183 | + panic("internal error") |
| 184 | + } else { |
| 185 | + if ok { |
| 186 | + m++ |
| 187 | + } |
| 188 | + } |
| 189 | + } |
| 190 | + return m |
| 191 | +} |
0 commit comments