prosite_search.cpp

来自「ncbi源码」· C++ 代码 · 共 374 行

/*
 * ===========================================================================
 * PRODUCTION $Log: prosite_search.cpp,v $
 * PRODUCTION Revision 1000.6  2004/06/01 20:55:39  gouriano
 * PRODUCTION PRODUCTION: UPGRADED [GCC34_MSVC7] Dev-tree R1.23
 * PRODUCTION
 * ===========================================================================
 */

/*  $Id: prosite_search.cpp,v 1000.6 2004/06/01 20:55:39 gouriano Exp $
 * ===========================================================================
 *
 *                            PUBLIC DOMAIN NOTICE
 *               National Center for Biotechnology Information
 *
 *  This software/database is a "United States Government Work" under the
 *  terms of the United States Copyright Act.  It was written as part of
 *  the author's official duties as a United States Government employee and
 *  thus cannot be copyrighted.  This software/database is freely available
 *  to the public for use. The National Library of Medicine and the U.S.
 *  Government have not placed any restriction on its use or reproduction.
 *
 *  Although all reasonable efforts have been taken to ensure the accuracy
 *  and reliability of the software and data, the NLM and the U.S.
 *  Government do not and cannot warrant the performance or results that
 *  may be obtained by using this software or data. The NLM and the U.S.
 *  Government disclaim all warranties, express or implied, including
 *  warranties of performance, merchantability or fitness for any particular
 *  purpose.
 *
 *  Please cite the author in any work or product based on this material.
 *
 * ===========================================================================
 *
 * Authors:  Josh Cherry
 *
 * File Description:  gbench plugin for searching against prosite patterns
 *
 */


#include <ncbi_pch.hpp>
#include "prosite_search.hpp"
#include "prosite.hpp"

#include <algo/sequence/find_pattern.hpp>
#include <corelib/ncbiapp.hpp>
#include <corelib/ncbireg.hpp>
#include <gui/core/plugin_utils.hpp>
#include <gui/utils/system_path.hpp>
#include <gui/core/version.hpp>
#include <gui/dialogs/col/multi_col_dlg.hpp>
#include <gui/plugin/PluginCommandSet.hpp>
#include <gui/plugin/PluginInfo.hpp>
#include <gui/plugin/PluginRequest.hpp>
#include <gui/plugin/PluginValueConstraint.hpp>
#include <gui/utils/message_box.hpp>
#include <gui/objutils/utils.hpp>
#include <objects/seqloc/Seq_interval.hpp>
#include <objmgr/seq_vector.hpp>
#include <objmgr/util/sequence.hpp>


BEGIN_NCBI_SCOPE
USING_SCOPE(objects);



CAlgoPlugin_PrositeSearch::~CAlgoPlugin_PrositeSearch()
{
}


// standard plugin announce bopilerplate
void CAlgoPlugin_PrositeSearch::GetInfo(CPluginInfo& info)
{
    info.Reset();
    
    // version info macro
    info.SetInfo(CPluginVersion::eMajor, CPluginVersion::eMinor, 0,
                 string(__DATE__) + " " + string(__TIME__),
                 "CAlgoPlugin_PrositeSearch", "Search/Search against prosite",
                 "Search a protein sequence against prosite patterns",
                 "");

    // command info
    CPluginCommandSet& cmds = info.SetCommands();
    CPluginCommand&    args = cmds.AddAlgoCommand(eAlgoCommand_run);
    args.AddArgument("locs", "Locations to evaluate",
                     CSeq_loc::GetTypeInfo(),
                     CPluginArg::TData::e_Array);
    args.SetConstraint("locs",
                       (*CPluginValueConstraint::CreateSeqMol(),
                        CSeq_inst::eMol_aa));
    args.AddDefaultArgument("skip",
                            "Skip patterns with SKIP_FLAG set",
                            CPluginArg::eBoolean, "true");
}


void CAlgoPlugin_PrositeSearch::RunCommand(CPluginMessage& msg)
{
    const CPluginCommand& args = msg.GetRequest().GetCommand();
    CPluginReply& reply = msg.SetReply();
    _TRACE("CAlgoPlugin_PrositeSearch::RunCommand()");

    bool skip = args["skip"].AsBoolean();
    
    // load patterns from file
    vector<CPrositeEntry> entries;
    try {
        x_LoadPrositeData(entries);
    }
    catch (const exception& e) {
        NcbiMessageBox(e.what());
        reply.SetStatus(eMessageStatus_failed);
        return;
    }


    if ( !m_Dialog.get() ) {
        m_Dialog.reset(new CMultiColDlg());
        m_Dialog->SetWindowSize(1000, 350);
        m_Dialog->SetTitle("Pattern Search Results");

        m_Dialog->SetColumn(0, "Sequence", FL_ALIGN_LEFT, 1.0f);
        m_Dialog->SetColumn(1, "Location", FL_ALIGN_LEFT, 1.0f);
        m_Dialog->SetColumn(2, "Position", FL_ALIGN_CENTER, 1.0f);
        m_Dialog->SetColumn(3, "Pattern ID", FL_ALIGN_LEFT, 3.0f);
        m_Dialog->SetColumn(4, "Pattern Description", FL_ALIGN_LEFT, 4.0f);
        m_Dialog->SetColumn(5, "Pattern", FL_ALIGN_LEFT, 3.0f);
        m_Dialog->SetColumn(6, "Matched Sequence", FL_ALIGN_LEFT, 2.0f);
    }

    m_Dialog->SetRows(0);  // to clear any previous contents
 
    vector<TSeqPos> starts;
    vector<TSeqPos> ends;

    //
    // first, evaluate whole sequences
    //
    int row = 0;
    plugin_args::TLocList locs;
    GetArgValue(args["locs"], locs);

    unsigned int patterns_searched;

    ITERATE (plugin_args::TLocList, iter, locs) {
        const CSeq_loc&  loc = *iter->second;
        const IDocument& doc = *iter->first;

        // find the best ID for this bioseq
        try {
            CBioseq_Handle handle = doc.GetScope().GetBioseqHandle(loc);
            CSeqVector vec =
                handle.GetSequenceView(loc,
                                       CBioseq_Handle::eViewConstructed,
                                       CBioseq_Handle::eCoding_Iupac);
            
            string seq;
            vec.GetSeqData( (TSeqPos) 0, vec.size(), seq );

            string& id_str  = m_Dialog->SetCell(row, 0);
            string& loc_str = m_Dialog->SetCell(row, 1);

            const CSeq_id& best_id =
                sequence::GetId(handle, sequence::eGetId_Best);
            id_str.erase();
            best_id.GetLabel(&id_str);
            loc_str = CPluginUtils::GetLabel(loc, &doc.GetScope());

            // a new feature table
            CRef<CSeq_annot> annot(new CSeq_annot());

            patterns_searched = 0;
            // iterate over patterns
            ITERATE (vector<CPrositeEntry>, entry, entries) {
                if (entry->GetSkipFlag() && skip) {
                    continue;  // skip this pattern entry
                }

                string pcre = CProsite::PSPatternToPCRE(entry->GetPattern());
                CFindPattern::Find(seq, pcre, starts, ends);
                patterns_searched++;

                for( unsigned int k = 0;  k < starts.size();  k++) {
                    string& pos_str = m_Dialog->SetCell(row, 2);
                    // 1-based indexing for dialog
                    pos_str = NStr::IntToString(starts[k] + 1) + " - "
                        + NStr::IntToString(ends[k] + 1);

                    m_Dialog->SetCell(row, 3) = entry->GetId();
                    m_Dialog->SetCell(row, 4) = entry->GetDesc();
                    m_Dialog->SetCell(row, 5) = entry->GetPattern();

                    m_Dialog->SetCell(row, 6) =
                        seq.substr(starts[k], ends[k] - starts[k] + 1);
                    ++row;
                }

                //
                // add features to annot
                //

                for( unsigned int k = 0;  k < starts.size();  k++) {
                    // create feature
                    CRef<CSeq_feat> feat(new CSeq_feat());

                    // set correct location
                    CSeq_loc& floc = feat->SetLocation();
                    floc.SetInt().SetId().Assign(sequence::GetId(loc));
                    floc.SetInt().SetFrom(starts[k]);
                    floc.SetInt().SetTo  (ends[k]);
                    feat->SetLocation
                        (*CSeqUtils::RemapChildToParent(loc, floc));

                    // set feature data
                    feat->SetData().SetRegion() = "Prosite match: "
                        + entry->GetDesc() + "; "
                        + entry->GetId();
    
                    // save in annot
                    annot->SetData().SetFtable().push_back(feat);
                }
            }

            // add description to annot
            annot->SetName("Prosite matches");

            // attach annot to doc
            //const_cast<IDocument&>(doc).AttachAnnot(*annot);
            reply.AddObject(doc, *annot);
        }
        catch (exception& e) {
            LOG_POST(Error << e.what());
            string str = CPluginUtils::GetLabel(loc, &doc.GetScope());
            LOG_POST(Error << "Error processing location " << str);
        }
#ifndef _DEBUG
        catch (...) {
            string str = CPluginUtils::GetLabel(loc, &doc.GetScope());
            LOG_POST(Error << "Error processing location " << str);
        }
#endif
    }

    // update all views
    //CDocManager::UpdateAllViews();

    //
    // prepare our dialog box
    //
    m_Dialog->SetLabel(string("A search against ")
                 + NStr::IntToString(patterns_searched)
                 + " patterns produced:");
    m_Dialog->Show();
    reply.SetStatus(eMessageStatus_success);
    reply.AddAction(CPluginReplyAction::e_Add_to_document);
}


void CAlgoPlugin_PrositeSearch::x_LoadPrositeData(vector<CPrositeEntry>& entries)
{
    CNcbiApplication* app = CNcbiApplication::Instance();
    _ASSERT(app);
    CNcbiRegistry& registry = app->GetConfig();
    string fname;

    // By default the rebase "NAR format" file 
    // is assumed to be <std>/etc/rebase.nar.
    // This can be overridden in gbench.ini, via the application registry
    // variable [PROSITE_SEARCH] PrositeData.
    fname = registry.GetString("PROSITE_SEARCH", "PrositeData", "");
    if ( !fname.empty() ) {
        fname += ", ";
    }
    fname += "<home>/etc/prosite.dat, <std>/etc/prosite.dat";
    fname = CSystemPath::ResolvePathExisting(fname);
    if ( fname.empty() ) {
        throw runtime_error("Couldn't open PROSITE file");
    }
    
    ifstream psfile(fname.c_str());
    CProsite::ReadEntries(psfile, entries);
}


END_NCBI_SCOPE


/*
 * ===========================================================================
 * $Log: prosite_search.cpp,v $
 * Revision 1000.6  2004/06/01 20:55:39  gouriano
 * PRODUCTION: UPGRADED [GCC34_MSVC7] Dev-tree R1.23
 *
 * Revision 1.23  2004/05/21 22:27:47  gorelenk
 * Added PCH ncbi_pch.hpp
 *
 * Revision 1.22  2004/05/03 13:05:42  dicuccio
 * gui/utils --> gui/objutils where needed
 *
 * Revision 1.21  2004/03/05 17:35:37  dicuccio
 * Use sequence::GetId() instead of CSeq_id::GetStringDescr()
 *
 * Revision 1.20  2004/02/17 20:35:25  rsmith
 * moved core/settings.[ch]pp and core/system_path.[ch]pp to config and utils, respectively.
 *
 * Revision 1.19  2004/02/13 15:08:50  mjohnson
 * Removed local URL for plugin help.
 *
 * Revision 1.18  2004/01/27 18:38:08  dicuccio
 * Code clean-up.  Use standard names for plugins.  Removed unnecessary #includes
 *
 * Revision 1.17  2004/01/07 15:50:38  dicuccio
 * Adjusted for API change in CPluginUtils::GetLabel().  Standardized exception
 * reporting in algorithms.
 *
 * Revision 1.16  2003/12/15 20:16:08  jcherry
 * Changed CFindPattern::Find to take a string rather than a CSeqVector
 *
 * Revision 1.15  2003/11/24 15:45:28  dicuccio
 * Renamed CVersion to CPluginVersion
 *
 * Revision 1.14  2003/11/18 17:48:38  dicuccio
 * Added standard processing of return values
 *
 * Revision 1.13  2003/11/14 00:20:29  jcherry
 * Added url for help
 *
 * Revision 1.12  2003/11/06 20:12:12  dicuccio
 * Cleaned up handling of USING_SCOPE - removed from all headers
 *
 * Revision 1.11  2003/11/04 17:49:23  dicuccio
 * Changed calling parameters for plugins - pass CPluginMessage instead of paired
 * CPluginCommand/CPluginReply
 *
 * Revision 1.10  2003/10/27 17:46:49  dicuccio
 * Removed dead #includes
 *
 * Revision 1.9  2003/10/14 16:24:02  dicuccio
 * Added correct remapping of scanned locations to the parent location.  Cleaned
 * up code to look for data file - added hierarchichal search through path in INI,
 * user's home directory, and finally system installed path.
 *
 * Revision 1.8  2003/10/07 13:47:00  dicuccio
 * Renamed CPluginURL* to CPluginValue*
 *
 * Revision 1.7  2003/09/25 17:21:35  jcherry
 * Added name to annot
 *
 * Revision 1.6  2003/09/04 14:05:24  dicuccio
 * Use IDocument instead of CDocument
 *
 * Revision 1.5  2003/09/03 14:46:53  rsmith
 * change namespace name from args to plugin_args to avoid clashes with variable names.
 *
 * Revision 1.4  2003/08/21 12:03:07  dicuccio
 * Make use of new typedef in plugin_utils.hpp for argument values.
 *
 * Revision 1.3  2003/08/15 18:48:06  jcherry
 * Brought dialog handling up to date.
 *
 * Revision 1.2  2003/08/11 18:03:25  jcherry
 * Fixed to use 0-based indexing for feature table (but retain
 * 1-based indexing for dialog box)
 *
 * Revision 1.1  2003/08/04 20:05:45  jcherry
 * Initial version
 *
 * ===========================================================================
 */