1 #include "mltaln.h"
2
3 #define DEBUG 0
4 #define IODEBUG 0
5
arguments(int argc,char * argv[])6 void arguments( int argc, char *argv[] )
7 {
8 int c;
9
10 calledByXced = 0;
11 devide = 0;
12 use_fft = 0;
13 fftscore = 1;
14 fftRepeatStop = 0;
15 fftNoAnchStop = 0;
16 weight = 3;
17 utree = 1;
18 tbutree = 1;
19 refine = 0;
20 check = 1;
21 cut = 0.0;
22 disp = 0;
23 outgap = 1;
24 alg = 'C';
25 mix = 0;
26 tbitr = 0;
27 scmtd = 5;
28 tbweight = 0;
29 tbrweight = 3;
30 checkC = 0;
31 treemethod = 'x';
32 contin = 0;
33 ppenalty = NOTSPECIFIED;
34 ppenalty_ex = NOTSPECIFIED;
35 poffset = NOTSPECIFIED;
36 kimuraR = NOTSPECIFIED;
37 pamN = NOTSPECIFIED;
38 geta2 = GETA2;
39 scoremtx = NOTSPECIFIED;
40
41 while( --argc > 0 && (*++argv)[0] == '-' )
42 {
43 while ( (c = *++argv[0]) )
44 {
45 switch( c )
46 {
47 case 'f':
48 ppenalty = (int)( atof( *++argv ) * 1000 - 0.5 );
49 fprintf( stderr, "ppenalty = %d\n", ppenalty );
50 --argc;
51 goto nextoption;
52 case 'g':
53 ppenalty_ex = (int)( atof( *++argv ) * 1000 - 0.5 );
54 fprintf( stderr, "ppenalty_ex = %d\n", ppenalty_ex );
55 --argc;
56 goto nextoption;
57 case 'h':
58 poffset = (int)( atof( *++argv ) * 1000 - 0.5 );
59 fprintf( stderr, "poffset = %d\n", poffset );
60 --argc;
61 goto nextoption;
62 case 'D':
63 scoremtx = -1;
64 break;
65 case 'P':
66 scoremtx = 0;
67 break;
68 case 'i':
69 contin = 1;
70 break;
71 case 'e':
72 fftscore = 0;
73 break;
74 case 'O':
75 fftNoAnchStop = 1;
76 break;
77 case 'R':
78 fftRepeatStop = 1;
79 break;
80 case 'Q':
81 calledByXced = 1;
82 break;
83 case 's':
84 treemethod = 's';
85 break;
86 case 'x':
87 treemethod = 'x';
88 break;
89 case 'p':
90 treemethod = 'p';
91 break;
92 case 'a':
93 alg = 'a';
94 break;
95 case 'A':
96 alg = 'A';
97 break;
98 case 'S':
99 alg = 'S';
100 break;
101 case 'C':
102 alg = 'C';
103 break;
104 case 'F':
105 use_fft = 1;
106 break;
107 case 'v':
108 tbrweight = 3;
109 break;
110 case 'd':
111 disp = 1;
112 break;
113 case 'o':
114 outgap = 0;
115 break;
116 /* Modified 01/08/27, default: user tree */
117 case 'J':
118 tbutree = 0;
119 break;
120 /* modification end. */
121 case 'Z':
122 checkC = 1;
123 break;
124 default:
125 fprintf( stderr, "illegal option %c\n", c );
126 argc = 0;
127 break;
128 }
129 }
130 nextoption:
131 ;
132 }
133 if( argc == 1 )
134 {
135 cut = atof( (*argv) );
136 argc--;
137 }
138 if( argc != 0 )
139 {
140 fprintf( stderr, "options: Check source file !\n" );
141 exit( 1 );
142 }
143 if( tbitr == 1 && outgap == 0 )
144 {
145 fprintf( stderr, "conflicting options : o, m or u\n" );
146 exit( 1 );
147 }
148 if( alg == 'C' && outgap == 0 )
149 {
150 fprintf( stderr, "conflicting options : C, o\n" );
151 exit( 1 );
152 }
153 readOtherOptions( &ppid, &fftThreshold, &fftWinSize );
154 }
155
156
treebase(char ** name,int nlen[M],char ** seq,char ** aseq,char ** mseq1,char ** mseq2,double ** mtx,int *** topol,double ** len,double ** eff,int alloclen)157 void treebase( char **name, int nlen[M], char **seq, char **aseq, char **mseq1, char **mseq2, double **mtx, int ***topol, double **len, double **eff, int alloclen )
158 {
159 int i, j, l;
160 int clus1, clus2;
161 int s1, s2, r1, r2;
162 double pscore;
163 static char *indication1, *indication2;
164 static char **name1, **name2;
165 static double **partialmtx = NULL;
166 static int ***partialtopol = NULL;
167 static double **partiallen = NULL;
168 static double **partialeff = NULL;
169 static double *effarr = NULL;
170 static double *effarr1 = NULL;
171 static double *effarr2 = NULL;
172 #if 0
173 char pair[njob][njob];
174 #else
175 static char **pair;
176 #endif
177 if( partialtopol == NULL )
178 {
179 partialmtx = AllocateDoubleMtx( njob, njob );
180 partialtopol = AllocateIntCub( njob, 2, njob );
181 partialeff = AllocateDoubleMtx( njob, njob );
182 partiallen = AllocateDoubleMtx( njob, 2 );
183 effarr = AllocateDoubleVec( njob );
184 effarr1 = AllocateDoubleVec( njob );
185 effarr2 = AllocateDoubleVec( njob );
186 indication1 = AllocateCharVec( njob*3+100 );
187 indication2 = AllocateCharVec( njob*3+100 );
188 name1 = AllocateCharMtx( njob, B+1 );
189 name2 = AllocateCharMtx( njob, B+1 );
190 #if 0
191 #else
192 pair = AllocateCharMtx( njob, njob );
193 #endif
194 }
195
196 if( checkC )
197 for( i=0; i<njob; i++ ) fprintf( stderr, "eff in tb-%d %f\n", i, eff[i][i] );
198
199 for( i=0; i<njob; i++ ) effarr[i] = eff[i][i];
200 for( i=0; i<njob; i++ ) strcpy( aseq[i], seq[i] );
201
202 if( checkC )
203 for( i=0; i<njob; i++ ) fprintf( stderr, "effarr for aseq-%d %f\n", i, effarr[i] );
204
205 writePre( njob, name, nlen, aseq, 0 );
206
207 for( i=0; i<njob; i++ ) for( j=0; j<njob; j++ ) pair[i][j] = 0;
208 for( i=0; i<njob; i++ ) pair[i][i] = 1;
209 for( l=0; l<njob-1; l++ )
210 {
211 s1 = topol[l][0][0];
212 for( i=0; (r1=topol[l][0][i])>-1; i++ )
213 if( pair[s1][r1] != 1 ) exit( 1 );
214 s2 = topol[l][1][0];
215 for( i=0; (r2=topol[l][1][i])>-1; i++ )
216 if( pair[s2][r2] != 1 ) exit( 1 );
217
218 clus1 = conjuction( pair, s1, aseq, mseq1, effarr1, effarr, name, name1, indication1 );
219 clus2 = conjuction( pair, s2, aseq, mseq2, effarr2, effarr, name, name2, indication2 );
220 fprintf( trap_g, "\nSTEP-%d\n", l );
221 fprintf( trap_g, "group1 = %s\n", indication1 );
222 fprintf( trap_g, "group2 = %s\n", indication2 );
223
224 fprintf( stderr, "STEP %d /%d\n", l+1, njob-1 );
225 fprintf( stderr, "group1 = %.66s", indication1 );
226 if( strlen( indication1 ) > 66 ) fprintf( stderr, "..." );
227 fprintf( stderr, "\n" );
228 fprintf( stderr, "group2 = %.66s", indication2 );
229 if( strlen( indication2 ) > 66 ) fprintf( stderr, "..." );
230 fprintf( stderr, "\n" );
231
232 if( checkC )
233 for( i=0; i<clus1; i++ ) fprintf( stderr, "STEP%d-eff for mseq1-%d %f\n", l+1, i, effarr1[i] );
234 /*
235 fprintf( stderr, "before align all\n" );
236 display( aseq, njob );
237 fprintf( stderr, "\n" );
238 fprintf( stderr, "before align 1 %s \n", indication1 );
239 display( mseq1, clus1 );
240 fprintf( stderr, "\n" );
241 fprintf( stderr, "before align 2 %s \n", indication2 );
242 display( mseq2, clus2 );
243 fprintf( stderr, "\n" );
244 */
245
246 pscore = Fgetlag( n_dis_consweight_multi, mseq1, mseq2, effarr1, effarr2, clus1, clus2, alloclen );
247
248 for( i=0; (r2=topol[l][1][i])>-1; i++ )
249 {
250 pair[s1][r2] = 1;
251 pair[s2][r2] = 0;
252 }
253
254 writePre( njob, name, nlen, aseq, 0 );
255
256 if( disp ) display( aseq, njob );
257 fprintf( stderr, "\n" );
258
259 }
260 }
261
WriteOptions(FILE * fp)262 static void WriteOptions( FILE *fp )
263 {
264 fprintf( fp, "tree-base method\n" );
265 if( tbrweight == 0 ) fprintf( fp, "unweighted\n" );
266 else if( tbrweight == 3 ) fprintf( fp, "clustalw-like weighting\n" );
267 if( tbitr || tbweight )
268 {
269 fprintf( fp, "iterate at each step\n" );
270 if( tbitr && tbrweight == 0 ) fprintf( fp, " unweighted\n" );
271 if( tbitr && tbrweight == 3 ) fprintf( fp, " reversely weighted\n" );
272 if( tbweight ) fprintf( fp, " weighted\n" );
273 fprintf( fp, "\n" );
274 }
275 if ( scoremtx == 0 ) fprintf( fp, "JTT %dPAM\n", pamN );
276 else if( scoremtx == 1 ) fprintf( fp, "Dayhoff( machigai ga aru )\n" );
277 else if( scoremtx == 2 ) fprintf( fp, "M-Y\n" );
278 else if( scoremtx == -1 ) fprintf( fp, "DNA\n" );
279
280 if( scoremtx == 0 || scoremtx == -1 )
281 fprintf( fp, "Gap Penalty = %+5.2f, %+5.2f, %+5.2f\n", (double)ppenalty/1000, (double)ppenalty_ex/1000, (double)poffset/1000 );
282 else
283 fprintf( fp, "Gap Penalty = %+5.2f\n", (double)ppenalty/1000 );
284
285 if( alg == 'a' )
286 fprintf( fp, "Algorithm A\n" );
287 else if( alg == 'A' )
288 fprintf( fp, "Apgorithm A+\n" );
289 else if( alg == 'S' )
290 fprintf( fp, "Apgorithm S\n" );
291 else if( alg == 'C' )
292 fprintf( fp, "Apgorithm A+/C\n" );
293 else
294 fprintf( fp, "Unknown algorithm\n" );
295
296 if( treemethod == 'x' )
297 fprintf( fp, "Tree = UPGMA (3).\n" );
298 else if( treemethod == 's' )
299 fprintf( fp, "Tree = UPGMA (2).\n" );
300 else if( treemethod == 'p' )
301 fprintf( fp, "Tree = UPGMA (1).\n" );
302 else
303 fprintf( fp, "Unknown tree.\n" );
304
305 if( use_fft )
306 {
307 fprintf( fp, "FFT on\n" );
308 if( scoremtx == -1 )
309 fprintf( fp, "Basis : 4 nucleotides\n" );
310 else
311 {
312 if( fftscore )
313 fprintf( fp, "Basis : Polarity and Volume\n" );
314 else
315 fprintf( fp, "Basis : 20 amino acids\n" );
316 }
317 fprintf( fp, "Threshold of anchors = %d%%\n", fftThreshold );
318 fprintf( fp, "window size of anchors = %dsites\n", fftWinSize );
319 }
320 else
321 fprintf( fp, "FFT off\n" );
322 fflush( fp );
323 }
324
325
main(int argc,char * argv[])326 int main( int argc, char *argv[] )
327 {
328 static int nlen[M];
329 static char **name, **seq;
330 static char **mseq1, **mseq2;
331 static char **aseq;
332 static char **bseq;
333 static double **pscore;
334 static double **eff;
335 static double **node0, **node1;
336 int i, j;
337 static int ***topol;
338 static double **len;
339 FILE *prep;
340 char c;
341 int alloclen;
342
343 arguments( argc, argv );
344 getnumlen( stdin );
345 rewind( stdin );
346
347 name = AllocateCharMtx( njob, B+1 );
348 seq = AllocateCharMtx( njob, nlenmax*5+1 );
349 aseq = AllocateCharMtx( njob, nlenmax*5+1 );
350 bseq = AllocateCharMtx( njob, nlenmax*5+1 );
351 mseq1 = AllocateCharMtx( njob, 0 );
352 mseq2 = AllocateCharMtx( njob, 0 );
353 alloclen = nlenmax*5;
354
355 topol = AllocateIntCub( njob, 2, njob );
356 len = AllocateDoubleMtx( njob, 2 );
357 pscore = AllocateDoubleMtx( njob, njob );
358 eff = AllocateDoubleMtx( njob, njob );
359 node0 = AllocateDoubleMtx( njob, njob );
360 node1 = AllocateDoubleMtx( njob, njob );
361
362 #if 0
363 Read( name, nlen, seq );
364 #else
365 readData_pointer( stdin, name, nlen, seq );
366 #endif
367
368 constants( njob, seq );
369
370 #if 0
371 fprintf( stderr, "params = %d, %d, %d\n", penalty, penalty_ex, offset );
372 #endif
373
374 initSignalSM();
375
376 initFiles();
377
378 WriteOptions( trap_g );
379
380 c = seqcheck( seq );
381 if( c )
382 {
383 fprintf( stderr, "Illeagal character %c\n", c );
384 exit( 1 );
385 }
386
387 writePre( njob, name, nlen, seq, 0 );
388
389 if( tbutree == 0 )
390 {
391 for( i=1; i<njob; i++ )
392 {
393 if( nlen[i] != nlen[0] )
394 {
395 fprintf( stderr, "Input pre-aligned seqences or make hat2.\n" );
396 exit( 1 );
397 }
398 }
399 for( i=0; i<njob-1; i++ ) for( j=i+1; j<njob; j++ )
400 {
401 /*
402 pscore[i][j] = (double)score_calc1( seq[i], seq[j] );
403 */
404 pscore[i][j] = (double)substitution_hosei( seq[i], seq[j] );
405 }
406 }
407 else
408 {
409 fprintf( stderr, "Loading 'hat2' ... " );
410 prep = fopen( "hat2", "r" );
411 if( prep == NULL ) ErrorExit( "Make hat2." );
412 readhat2_pointer( prep, njob, name, pscore );
413 fclose( prep );
414 fprintf( stderr, "done.\n" );
415
416 #if 0
417 prep = fopen( "hat2_check", "w" );
418 WriteHat2( prep, njob, name, pscore );
419 fclose( prep );
420 #endif
421
422 }
423
424 fprintf( stderr, "Constructing dendrogram ... " );
425 if( treemethod == 'x' )
426 supg( njob, pscore, topol, len );
427 else if( treemethod == 's' )
428 spg( njob, pscore, topol, len );
429 else if( treemethod == 'p' )
430 upg2( njob, pscore, topol, len );
431 else
432 ErrorExit( "Incorrect tree\n" );
433 fprintf( stderr, "done.\n" );
434
435 countnode( njob, topol, node0 );
436 if( tbrweight )
437 {
438 weight = 3;
439 utree = 0; counteff( njob, topol, len, eff ); utree = 1;
440 }
441 else
442 {
443 for( i=0; i<njob; i++ ) eff[i][i] = 1.0;
444 }
445
446
447 for( i=0; i<njob; i++ ) gappick0( bseq[i], seq[i] );
448
449 treebase( name, nlen, bseq, aseq, mseq1, mseq2, pscore, topol, len, eff, alloclen );
450
451 fprintf( trap_g, "done\n" );
452 fclose( trap_g );
453
454 writePre( njob, name, nlen, aseq, !contin );
455 writeData_pointer( stdout, njob, name, nlen, aseq );
456 #if IODEBUG
457 fprintf( stderr, "OSHIMAI\n" );
458 #endif
459 SHOWVERSION;
460 return( 0 );
461 }
462