1 |
duarte |
420 |
package proteinstructure; |
2 |
|
|
|
3 |
duarte |
452 |
import java.util.ArrayList; |
4 |
|
|
import java.util.Collection; |
5 |
stehr |
515 |
import java.util.LinkedList; |
6 |
duarte |
420 |
import java.util.TreeMap; |
7 |
|
|
|
8 |
|
|
|
9 |
stehr |
515 |
/** |
10 |
|
|
* A neighbourhood in a RIGraph, i.e. a collection of nodes connected to a 'central' node. |
11 |
|
|
* Note that the collection of neighbours now also contains the central residue. |
12 |
|
|
* To get strictly the neighbours, use the getNeighbors() method or grab them directly |
13 |
|
|
* from the original graph using getNeighbors(). |
14 |
|
|
*/ |
15 |
duarte |
420 |
public class RIGNbhood extends TreeMap<Integer,RIGNode> { |
16 |
|
|
|
17 |
|
|
|
18 |
|
|
private static final long serialVersionUID = 1L; |
19 |
|
|
|
20 |
|
|
public static final String centralLetter="x"; |
21 |
duarte |
452 |
public static final String gapLetter="_"; |
22 |
duarte |
420 |
|
23 |
|
|
// central residue |
24 |
|
|
private RIGNode centralResidue; |
25 |
|
|
|
26 |
|
|
/** |
27 |
|
|
* Construct a RIGNbhood by passing the central residue as a RIGNode object |
28 |
duarte |
452 |
* and all neighbors as a Collection<RIGNode> |
29 |
duarte |
420 |
* @param centralResidue |
30 |
duarte |
452 |
* @param neighbors |
31 |
duarte |
420 |
*/ |
32 |
duarte |
452 |
public RIGNbhood(RIGNode centralResidue, Collection<RIGNode> neighbors){ |
33 |
duarte |
420 |
super(); |
34 |
|
|
this.centralResidue = centralResidue; |
35 |
duarte |
452 |
this.put(centralResidue.getResidueSerial(), centralResidue); |
36 |
|
|
for (RIGNode node:neighbors) { |
37 |
|
|
this.put(node.getResidueSerial(),node); |
38 |
|
|
} |
39 |
duarte |
420 |
} |
40 |
duarte |
452 |
|
41 |
|
|
/** |
42 |
|
|
* Constructs a RIGNbhood with just a central residue and no neighbors |
43 |
|
|
* @param centralResidue |
44 |
|
|
*/ |
45 |
|
|
public RIGNbhood(RIGNode centralResidue) { |
46 |
|
|
super(); |
47 |
|
|
this.centralResidue = centralResidue; |
48 |
|
|
this.put(centralResidue.getResidueSerial(), centralResidue); |
49 |
|
|
} |
50 |
duarte |
420 |
|
51 |
|
|
public RIGNode getCentralResidue() { |
52 |
|
|
return centralResidue; |
53 |
|
|
} |
54 |
|
|
|
55 |
duarte |
452 |
/** |
56 |
|
|
* Returns true if one of the neighbors or the central residue is of the given residue type |
57 |
|
|
* @param resType |
58 |
|
|
* @return |
59 |
|
|
*/ |
60 |
|
|
public boolean containsResType(String resType) { |
61 |
|
|
for (RIGNode node:this.values()) { |
62 |
|
|
if (node.getResidueType().equals(resType)) { |
63 |
|
|
return true; |
64 |
|
|
} |
65 |
|
|
} |
66 |
|
|
return false; |
67 |
|
|
} |
68 |
|
|
|
69 |
|
|
/** |
70 |
|
|
* Returns whether this RIGNbhood is equal to the given one |
71 |
|
|
* Equality defined as: same residue types in the same order, both to the left and right of central residue |
72 |
|
|
* @param other |
73 |
|
|
* @return |
74 |
|
|
*/ |
75 |
|
|
public boolean equals(Object other) { |
76 |
|
|
|
77 |
|
|
|
78 |
|
|
RIGNbhood otherNbhood = (RIGNbhood) other; |
79 |
|
|
if (this.size()!=otherNbhood.size()) { |
80 |
|
|
return false; |
81 |
|
|
} |
82 |
|
|
if (!this.centralResidue.getResidueType().equals(otherNbhood.centralResidue.getResidueType())) { |
83 |
|
|
return false; |
84 |
|
|
} |
85 |
|
|
|
86 |
|
|
// we order the RIGNodes in 2 groups: left and right of the central residues (both for this and other) |
87 |
|
|
ArrayList<RIGNode> thisLeft = new ArrayList<RIGNode>(); |
88 |
|
|
ArrayList<RIGNode> thisRight = new ArrayList<RIGNode>(); |
89 |
|
|
ArrayList<RIGNode> otherLeft = new ArrayList<RIGNode>(); |
90 |
|
|
ArrayList<RIGNode> otherRight = new ArrayList<RIGNode>(); |
91 |
|
|
for (RIGNode node:this.values()) { |
92 |
|
|
if (node.getResidueSerial()<this.centralResidue.getResidueSerial()) { |
93 |
|
|
thisLeft.add(node); |
94 |
|
|
} else if (node.getResidueSerial()!=this.centralResidue.getResidueSerial()){ |
95 |
|
|
thisRight.add(node); |
96 |
|
|
} |
97 |
|
|
} |
98 |
|
|
for (RIGNode node:otherNbhood.values()) { |
99 |
|
|
if (node.getResidueSerial()<otherNbhood.centralResidue.getResidueSerial()) { |
100 |
|
|
otherLeft.add(node); |
101 |
|
|
} else if (node.getResidueSerial()!=otherNbhood.centralResidue.getResidueSerial()){ |
102 |
|
|
otherRight.add(node); |
103 |
|
|
} |
104 |
|
|
} |
105 |
|
|
|
106 |
|
|
// if sizes don't match they are different |
107 |
|
|
if (thisLeft.size()!=otherLeft.size()) { |
108 |
|
|
return false; |
109 |
|
|
} |
110 |
|
|
if (thisRight.size()!=otherRight.size()) { |
111 |
|
|
return false; |
112 |
|
|
} |
113 |
|
|
|
114 |
|
|
// as the RIGNodes are ordered as originally (sequence order) we simply compare if types match (left and then right) |
115 |
|
|
for (int i=0;i<thisLeft.size();i++) { |
116 |
|
|
if (!thisLeft.get(i).getResidueType().equals(otherLeft.get(i).getResidueType())) { |
117 |
|
|
return false; |
118 |
|
|
} |
119 |
|
|
} |
120 |
|
|
for (int i=0;i<thisRight.size();i++) { |
121 |
|
|
if (!thisRight.get(i).getResidueType().equals(otherRight.get(i).getResidueType())) { |
122 |
|
|
return false; |
123 |
|
|
} |
124 |
|
|
} |
125 |
|
|
|
126 |
|
|
// if we are here, we passed all tests: they are equal! |
127 |
|
|
return true; |
128 |
|
|
} |
129 |
|
|
|
130 |
|
|
|
131 |
|
|
/** |
132 |
|
|
* Returns true if this neighborhood matches the given one. |
133 |
|
|
* A match is defined as: both left and right sides of the this central |
134 |
|
|
* residue are substrings of other's left and right sides. |
135 |
|
|
* e.g.: |
136 |
|
|
* this: ABxEF matches: |
137 |
|
|
* other: --A-----B---xE---F--- (with '-' being other residues in between) |
138 |
|
|
* @param other |
139 |
|
|
* @return |
140 |
|
|
*/ |
141 |
|
|
public boolean match(RIGNbhood other) { |
142 |
|
|
// we order the RIGNodes in 2 groups: left and right of the central residues (both for this and other) |
143 |
|
|
ArrayList<RIGNode> thisLeft = new ArrayList<RIGNode>(); |
144 |
|
|
ArrayList<RIGNode> thisRight = new ArrayList<RIGNode>(); |
145 |
|
|
ArrayList<RIGNode> otherLeft = new ArrayList<RIGNode>(); |
146 |
|
|
ArrayList<RIGNode> otherRight = new ArrayList<RIGNode>(); |
147 |
|
|
for (RIGNode node:this.values()) { |
148 |
|
|
if (node.getResidueSerial()<this.centralResidue.getResidueSerial()) { |
149 |
|
|
thisLeft.add(node); |
150 |
|
|
} else if (node.getResidueSerial()!=this.centralResidue.getResidueSerial()){ |
151 |
|
|
thisRight.add(node); |
152 |
|
|
} |
153 |
|
|
} |
154 |
|
|
for (RIGNode node:other.values()) { |
155 |
|
|
if (node.getResidueSerial()<other.centralResidue.getResidueSerial()) { |
156 |
|
|
otherLeft.add(node); |
157 |
|
|
} else if (node.getResidueSerial()!=other.centralResidue.getResidueSerial()){ |
158 |
|
|
otherRight.add(node); |
159 |
|
|
} |
160 |
|
|
} |
161 |
|
|
|
162 |
|
|
// as the RIGNodes are ordered as originally (sequence order) we simply compare if types match (left and then right) |
163 |
|
|
int i = 0; |
164 |
|
|
int j = 0; |
165 |
|
|
while (i<thisLeft.size()) { |
166 |
|
|
if (j==otherLeft.size()) { |
167 |
|
|
return false; |
168 |
|
|
} |
169 |
|
|
if (thisLeft.get(i).getResidueType().equals(otherLeft.get(j).getResidueType())) { |
170 |
|
|
i++; |
171 |
|
|
j++; |
172 |
|
|
} else { |
173 |
|
|
j++; |
174 |
|
|
} |
175 |
|
|
} |
176 |
|
|
|
177 |
|
|
i = 0; |
178 |
|
|
j = 0; |
179 |
|
|
while (i<thisRight.size()) { |
180 |
|
|
if (j==otherRight.size()) { |
181 |
|
|
return false; |
182 |
|
|
} |
183 |
|
|
if (thisRight.get(i).getResidueType().equals(otherRight.get(j).getResidueType())) { |
184 |
|
|
i++; |
185 |
|
|
j++; |
186 |
|
|
} else { |
187 |
|
|
j++; |
188 |
|
|
} |
189 |
|
|
} |
190 |
|
|
|
191 |
|
|
// if we are here, we passed all tests: they match! |
192 |
|
|
return true; |
193 |
|
|
} |
194 |
|
|
|
195 |
duarte |
420 |
public String getMotifFullGaps(){ |
196 |
|
|
String motif=""; |
197 |
duarte |
452 |
for (int i=this.firstKey();i<=this.lastKey();i++) { |
198 |
|
|
if (this.containsKey(i)){ |
199 |
|
|
if (i!=centralResidue.getResidueSerial()){ |
200 |
duarte |
420 |
motif+=AAinfo.threeletter2oneletter(this.get(i).getResidueType()); |
201 |
duarte |
452 |
} else { |
202 |
duarte |
420 |
motif+=centralLetter; |
203 |
|
|
} |
204 |
duarte |
452 |
} else { |
205 |
|
|
motif+=gapLetter; |
206 |
duarte |
420 |
} |
207 |
|
|
} |
208 |
|
|
return motif; |
209 |
|
|
} |
210 |
|
|
|
211 |
|
|
public String getMotif(){ |
212 |
|
|
String motif=""; |
213 |
duarte |
452 |
int gapSize = 0; |
214 |
|
|
String gap = ""; |
215 |
|
|
for (int i=this.firstKey();i<=this.lastKey();i++) { |
216 |
|
|
if (this.containsKey(i)){ |
217 |
|
|
if (i!=centralResidue.getResidueSerial()){ |
218 |
duarte |
420 |
motif+=gap; |
219 |
|
|
motif+=AAinfo.threeletter2oneletter(this.get(i).getResidueType()); |
220 |
|
|
gapSize=0; |
221 |
|
|
gap=""; |
222 |
duarte |
452 |
} else { |
223 |
duarte |
420 |
motif+=gap; |
224 |
|
|
motif+=centralLetter; |
225 |
|
|
gapSize=0; |
226 |
|
|
gap=""; |
227 |
|
|
} |
228 |
duarte |
452 |
} else { |
229 |
|
|
gapSize++; |
230 |
|
|
gap=gapLetter+"{"+gapSize+"}"; |
231 |
duarte |
420 |
} |
232 |
|
|
} |
233 |
|
|
return motif; |
234 |
|
|
} |
235 |
|
|
|
236 |
|
|
public String getMotifNoGaps(){ |
237 |
|
|
String motif=""; |
238 |
duarte |
452 |
for (int i:this.keySet()) { |
239 |
|
|
if (i!=centralResidue.getResidueSerial()){ |
240 |
|
|
motif+=AAinfo.threeletter2oneletter(this.get(i).getResidueType()); |
241 |
|
|
} else { |
242 |
|
|
motif+=centralLetter; |
243 |
duarte |
420 |
} |
244 |
|
|
} |
245 |
|
|
return motif; |
246 |
|
|
} |
247 |
|
|
|
248 |
|
|
public String getMotifReducedAlphabet(RIGraph graph) { |
249 |
|
|
String motif=""; |
250 |
duarte |
452 |
for (int i:this.keySet()) { |
251 |
|
|
if (i!=centralResidue.getResidueSerial()){ |
252 |
|
|
// as long as the graph is undirected, JUNG finds the edge correctly independently of the order in which we give the nodes |
253 |
|
|
RIGEdge edge = graph.findEdge(centralResidue,this.get(i)); |
254 |
|
|
double weight = edge.getWeight(); |
255 |
|
|
if (weight>0) { // SC dominated |
256 |
|
|
motif+=AAinfo.threeletter2oneletter(this.get(i).getResidueType()); |
257 |
|
|
} else if (weight<0) { //BB dominated |
258 |
|
|
// for the sec structure we take the "seen" node (so not the central but the other), following Michael's convention |
259 |
|
|
char ssType = this.get(i).getSecStrucElement().getType(); |
260 |
|
|
char ssLetter = 'o'; |
261 |
|
|
if (ssType == 0 || ssType == 'O') ssLetter='o'; |
262 |
|
|
if (ssType=='S') ssLetter = 'b'; |
263 |
|
|
else if (ssType=='H') ssLetter= 'z'; |
264 |
|
|
motif+=ssLetter; |
265 |
duarte |
420 |
} |
266 |
duarte |
452 |
} else { |
267 |
|
|
motif+=centralLetter; |
268 |
duarte |
420 |
} |
269 |
|
|
} |
270 |
duarte |
452 |
return motif; |
271 |
duarte |
420 |
} |
272 |
|
|
|
273 |
|
|
public String getCommaSeparatedResSerials(){ |
274 |
|
|
String ressers=""; |
275 |
|
|
for (int resser:this.keySet()){ |
276 |
duarte |
452 |
if (resser!=centralResidue.getResidueSerial()) { |
277 |
duarte |
454 |
ressers += resser+","; |
278 |
duarte |
452 |
} |
279 |
duarte |
420 |
} |
280 |
duarte |
454 |
// we chop off the last comma |
281 |
|
|
if (ressers.length()!=0) { |
282 |
|
|
ressers = ressers.substring(0, ressers.length()-1); |
283 |
|
|
} |
284 |
duarte |
420 |
return ressers; |
285 |
|
|
} |
286 |
|
|
|
287 |
|
|
public String toString(){ |
288 |
|
|
if (this.isEmpty()) return ""; |
289 |
|
|
else return this.getMotif(); |
290 |
|
|
} |
291 |
|
|
|
292 |
|
|
/** |
293 |
stehr |
515 |
* Returns a collection of the neighbours in this RIGNbh. |
294 |
|
|
* @return A (possibly empty) collection of RIGNodes. |
295 |
|
|
*/ |
296 |
|
|
public Collection<RIGNode> getNeighbors() { |
297 |
|
|
LinkedList<RIGNode> nbs = new LinkedList<RIGNode>(); |
298 |
|
|
for(RIGNode n:this.values()) { |
299 |
|
|
if(n != centralResidue) { |
300 |
|
|
nbs.add(n); |
301 |
|
|
} |
302 |
|
|
} |
303 |
|
|
return nbs; |
304 |
|
|
} |
305 |
|
|
|
306 |
|
|
/** |
307 |
duarte |
452 |
* Returns a copy (deep) of this RIGNbhood |
308 |
duarte |
420 |
* @return |
309 |
|
|
*/ |
310 |
|
|
public RIGNbhood copy(){ |
311 |
duarte |
452 |
RIGNbhood copy = new RIGNbhood(centralResidue.copy()); |
312 |
|
|
for (int residueSerial:this.keySet()) { |
313 |
|
|
if (residueSerial!=centralResidue.getResidueSerial()) { |
314 |
|
|
copy.put(residueSerial, this.get(residueSerial).copy()); |
315 |
|
|
} |
316 |
duarte |
420 |
} |
317 |
|
|
return copy; |
318 |
|
|
} |
319 |
|
|
} |