--- /dev/null
+/* -*- mode: c; tab-width: 4; c-basic-offset: 4; indent-tabs-mode: nil -*- */
+
+/*********************************************************************
+ * Clustal Omega - Multiple sequence alignment
+ *
+ * Copyright (C) 2010 University College Dublin
+ *
+ * Clustal-Omega is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License as
+ * published by the Free Software Foundation; either version 2 of the
+ * License, or (at your option) any later version.
+ *
+ * This file is part of Clustal-Omega.
+ *
+ ********************************************************************/
+
+/*
+ * RCS $Id: hhalignment.h 154 2010-11-09 18:29:05Z fabian $
+ */
+
+// hhalignment.h
+
+class Alignment
+{
+public:
+ int L; // number of match states of alignment
+ int N_in; // total number of sequences in alignment
+ int N_filtered; /* number of sequences after sequence identity
+ filtering */
+ int N_ss; // number of >ss_ or >sa sequences
+
+ int kss_dssp; /* index of sequence with secondary structure
+ by dssp -1:no >ss_dssp line found */
+ int ksa_dssp; /* index of sequence with solvent accessibility
+ by dssp -1:no >sa_dssp line found */
+ int kss_pred; /* index of sequence with predicted secondary
+ structure -1:no >ss_pred line found */
+ int kss_conf; /* index of sequence with confidence values of
+ prediction -1:no >ss_conf line found */
+ int kfirst; // index of first real sequence
+
+ char* longname; /* Full name of first sequence of original alignment
+ (NAME field) */
+ char name[NAMELEN]; // HMM name = first word in longname in lower case
+ char fam[NAMELEN]; // family ID (derived from name) (FAM field)
+ char file[NAMELEN]; /* Rootname (w/o path, with extension) of alignment
+ file that is used to construct the HMM */
+
+ int n_display; /* number of sequences to be displayed
+ (INCLUDING >ss_pred, >ss_conf, >ss_dssp sequences) */
+ char** sname; // names of display sequences (first seq=0, first char=0)
+ char** seq; // residues of display sequences (first char=1)
+ int* l; // l[i] = position of i'th match state in alignment
+
+ char* keep; /* keep[k]=1 if sequence is included in amino acid
+ frequencies; 0 otherwise (first=0) */
+
+ double *pdExWeight; /* external sequence weight as given by tree FIXME (FS) */
+
+ Alignment(int maxseq=MAXSEQ, int maxres=/*MAXRES*/par.maxResLen);
+ ~Alignment();
+
+ // Read alignment into X (uncompressed) in ASCII characters
+ void Read(FILE* inf, char infile[NAMELEN], char* line=NULL);
+#ifdef CLUSTALO
+ void Transfer(char **ppcProf, int iCnt);
+ void ClobberGlobal();
+#endif
+
+ /* Convert ASCII to numbers between 0 and 20, throw out all insert states,
+ record their number in I[k][i] and store sequences to be displayed
+ in sname[k] and seq[k] */
+ void Compress(const char infile[NAMELEN]);
+
+ // Apply sequence identity filter
+ inline int FilterForDisplay(int max_seqid, int coverage=0, int qid=0, float qsc=0, int N=0);
+ inline int Filter(int max_seqid, int coverage=0, int qid=0, float qsc=0, int N=0);
+ int Filter2(char keep[], int coverage, int qid, float qsc, int seqid1, int seqid2, int Ndiff);
+
+ bool FilterNeff(); /* MR1 */
+ float filter_by_qsc(float qsc, char* dummy); /* MR1 */
+
+ // Filter alignment for min score per column with core query profile, defined by min_coverage_core and min_seqid_core
+ int HomologyFilter(int coverage_core, float qsc_core, float coresc);
+
+ // Calculate AA frequencies q.p[i][a] and transition probabilities q.tr[i][a] from alignment
+ void FrequenciesAndTransitions(HMM& q, char* in=NULL);
+
+ // Calculate freqs q.f[i][a] and transitions q.tr[i][a] (a=MM,MI,MD) with pos-specific subalignments
+ void Amino_acid_frequencies_and_transitions_from_M_state(HMM& q, char* in);
+
+ // Calculate transitions q.tr[i][a] (a=DM,DD) with pos-specific subalignments
+ void Transitions_from_D_state(HMM& q, char* in);
+
+ // Calculate transitions q.tr[i][a] (a=DM,DD) with pos-specific subalignments
+ void Transitions_from_I_state(HMM& q, char* in);
+
+ // Write alignment without insert states to alignment file
+ void WriteWithoutInsertsToFile(char* alnfile);
+
+ // Write alignment to alignment file
+ void WriteToFile(char* alnfile, const char format[]=NULL);
+
+ // Read a3m slave alignment of hit from ta3mfile and merge into (query) master alignment
+ void MergeMasterSlave(Hit& hit, char ta3mfile[]);
+
+ // Read a3m alignment of hit from ta3mfile and merge-combine with query alignment
+ void Merge(Hit& hit, char ta3mfile[]);
+
+ // Add a sequence to Qali
+ void AddSequence(char Xk[], int Ik[]=NULL);
+
+ // Determine matrix of position-specific weights w[k][i] for multiple alignment
+ void GetPositionSpecificWeights(float* w[]);
+
+ char readCommentLine; // Set to 1, if a comment line with '#' is read /* MR1 */
+
+private:
+ char** X; // X[k][i] contains column i of sequence k in alignment (first seq=0, first char=1) (0-3: ARND ..., 20:X, 21:GAP)
+ short unsigned int** I; // I[k][i] contains the number of inserts AFTER match state i (first=0)
+ char* display; // display[k]=1 if sequence will be displayed in output alignments; 0 otherwise (first=0)
+ float* wg; // w[k] = global weight of sequence k
+ int* nseqs; // number of sequences in subalignment i (only for DEBUGGING)
+ int* nres; // number of residues in sequence k
+ int* first; // first residue in sequence k
+ int* last; // last residue in sequence k
+ int* ksort; // index for sorting sequences: X[ksort[k]]
+ int FilterWithCoreHMM(char in[], float coresc, HMM& qcore);
+};
git://source.jalview.org / jabaws.git / blobdiff