owl/trunk/calculateGridDensity.java

import java.io.*;
import java.sql.*;
import java.util.*;
import proteinstructure.*;
import tools.MySQLConnection;

// TODO:
// - how will the distribution look like for randomly drawing from the unit square? (use Ioannis' random graphs?)
// - what is the maximum and expected number of points per square? how does that translate into runtime
// Problems:
// - surface effect (distinguish surface/core)
// - translation of grid (-> kdtrees)
// Space of grid:
// - when assuming spherical proteins and constant density, grid size grows linearly
// - in the worst case of completely linear proteins, grid size growth is cubic
// - for the protein case, do benchmark (to prove linear runtime, estimate runtime constants and space exponent)
// - test whether rotating (by using principle components?) improves runtime/space

public class calculateGridDensity {

        /*------------------------------ constants ------------------------------*/
        
        // database with a list of pdb codes and chain codes to process
        public static String                    DB_NAME = "pdb_reps";
        public static String                    DB_TABLE = "reps";
        public static String                    DB_COL_PDB = "accession_code";
        public static String                    DB_COL_CHAIN = "chain_pdb_code";        
        
        public static String                    PDB_DB = "pdbase";
        public static String                    DB_HOST = "white";                                                              
        public static String                    DB_USER = getUserName();
        public static String                    DB_PWD = "nieve";

        public static String                    PDB_CODE = "1tdr";
        public static String                    CHAIN_CODE = "B";
        public static String                    edgeType = "Ca";
        public static double                    cutoff_from = 4.0;
        public static double                    cutoff_to = 15.0;
        public static double                    cutoff_step = 1.0;
        public static int                               limit = 100;
        
        public static String                    outFileName = "grid_nbs_pdbreps100";
        
        /*---------------------------- private methods --------------------------*/
        /** 
         * Get user name from operating system (for use as database username). 
         * */
        private static String getUserName() {
                String user = null;
                user = System.getProperty("user.name");
                if(user == null) {
                        System.err.println("Could not get user name from operating system.");
                }
                return user;
        }
        
        private static void calcDensity(String pdbCode, String chainCode, double cutoff, String egdeType, MySQLConnection conn, Map<Integer, Integer> densityCount) {
                Pdb pdb = null;
                try {
                        pdb = new PdbasePdb(pdbCode, chainCode, PDB_DB, conn);
                        // add to density count vector
                        pdb.calcGridDensity(edgeType, cutoff, densityCount);
                        
                } catch (PdbaseInconsistencyError e) {
                        System.out.println("Inconsistency in " + pdbCode + chainCode);
                } catch (PdbCodeNotFoundError e) {
                        e.printStackTrace();
                } catch (SQLException e) {
                        e.printStackTrace();
                } catch (PdbChainCodeNotFoundError e) {
                        e.printStackTrace();
                }
                        
        }
        
        public static void printValues(Map<Integer, Integer> v, PrintStream out) {
                int atoms = 0;
                for(int size:v.keySet()) {
                        out.println(size + "\t" + v.get(size));
                        atoms += size*v.get(size);
                }
                //out.println("Atoms: " + atoms);
        }
        
        
        public static void writeResultToFile(Map<Integer, Integer> v, String baseName, String edgeType, double cutoff) {
                try {
                        File outFile = new File(baseName + "_" + edgeType + "_" + cutoff + ".out");
                        if(outFile.exists()) {
                                outFile.delete();
                        }
                        outFile.createNewFile();
                        printValues(v, new PrintStream(outFile));
                        System.out.println("Results written to file " + outFile.getName());
                
                } catch (FileNotFoundException e) {
                        e.printStackTrace();
                } catch (IOException e) {
                        e.printStackTrace();
                }
        }
        
        
        public static void writeResultToDb(Map<Integer, Integer> v) {
                // insert into runs(run_id, edgeType, cutoff, timestamp, proteins, points, cells, avg_pts_per_cell, max_pts_per_cell, avg_pts_per_area, max_pts_per_area)
                // insert into density_distr(run_id, points_per_cell, num_cells)
                // do another run for values per protein (run_id, pdb_id, chain_id, num_res, num_atoms, num_cells, surface_cells, core_cells, surface_area, volume)

        }
        
        public static Map<Integer, Integer> calcDensity(MySQLConnection conn, String edgeType, double cutoff, boolean verbose) {
                Map<Integer, Integer> densityCount = new TreeMap<Integer,Integer>();
                String pdbCode, chainCode;
                int numPdbs = 0;
                
                if(verbose) {
                        System.out.print(edgeType + " " + cutoff + ": ");
                }
                
                // read structures from database
                String query = "SELECT DISTINCT " + DB_COL_PDB + "," + DB_COL_CHAIN + " FROM " + DB_NAME + "." + DB_TABLE + " LIMIT " + limit + ";" ;
                Statement stmt;
                try {
                        stmt = conn.createStatement();
                        ResultSet rs = stmt.executeQuery(query);
                        while(rs.next()) {
                                pdbCode = rs.getString(1);
                                chainCode = rs.getString(2);
                                
                                if(chainCode == null) {
                                        chainCode = "NULL";
                                }
                                numPdbs++;
                                
                                // calculate statistics
                                calcDensity(pdbCode, chainCode, cutoff, edgeType, conn, densityCount); // will add to densityCount
                                if(verbose) {
                                        if(numPdbs %2 == 0) {
                                                System.out.print('\b');
                                        } else {
                                                System.out.print(".");
                                        }
                                        if(numPdbs % 500 == 0) System.out.print(numPdbs + " ");
                                }
                                
                                // for each protein write to db: pdb, chain, num_res, volume, max_density
                        }
                        rs.close();
                        stmt.close();
                        if(verbose) System.out.println(".");
                        
                } catch (SQLException e) {
                        e.printStackTrace();
                }               
                return densityCount;
        }
        
        public static void main(String[] args) {
                MySQLConnection conn = null;
                Map<Integer, Integer> densityCount = null;
                double cutoff;

                // opening db connection
                try{
                        conn = new MySQLConnection(DB_HOST, DB_USER, DB_PWD);
                } catch (Exception e) {
                        System.err.println("Error opening database connection. Exiting");
                        System.exit(1);
                }

                for(cutoff = cutoff_from; cutoff <= cutoff_to; cutoff += cutoff_step) {
                        // run calculation
                        densityCount = calcDensity(conn, edgeType, cutoff, true);

                        // output results
                        //printValues(densityCount, System.out);
                        writeResultToFile(densityCount, outFileName, edgeType, cutoff);
                }               
                System.out.println("Done.");

        }

}
Revision:	419
Committed:	Thu Nov 22 14:09:18 2007 UTC (17 years, 3 months ago) by duarte
Original Path:	*branches/aglappe-jung/calculateGridDensity.java*
File size:	6000 byte(s)
Log Message:	Re-branching for JUNG2 development
Line	User	Rev	File contents
1	stehr	303	import java.io.*;
2	stehr	250	import java.sql.*;
3			import java.util.*;
4			import proteinstructure.*;
5			import tools.MySQLConnection;
6
7	stehr	303	// TODO:
8			// - how will the distribution look like for randomly drawing from the unit square? (use Ioannis' random graphs?)
9			// - what is the maximum and expected number of points per square? how does that translate into runtime
10			// Problems:
11			// - surface effect (distinguish surface/core)
12			// - translation of grid (-> kdtrees)
13			// Space of grid:
14			// - when assuming spherical proteins and constant density, grid size grows linearly
15			// - in the worst case of completely linear proteins, grid size growth is cubic
16			// - for the protein case, do benchmark (to prove linear runtime, estimate runtime constants and space exponent)
17			// - test whether rotating (by using principle components?) improves runtime/space
18	stehr	250
19			public class calculateGridDensity {
20
21			/------------------------------ constants ------------------------------/
22
23			// database with a list of pdb codes and chain codes to process
24			public static String DB_NAME = "pdb_reps";
25			public static String DB_TABLE = "reps";
26			public static String DB_COL_PDB = "accession_code";
27			public static String DB_COL_CHAIN = "chain_pdb_code";
28
29			public static String PDB_DB = "pdbase";
30			public static String DB_HOST = "white";
31			public static String DB_USER = getUserName();
32			public static String DB_PWD = "nieve";
33
34			public static String PDB_CODE = "1tdr";
35			public static String CHAIN_CODE = "B";
36			public static String edgeType = "Ca";
37	stehr	303	public static double cutoff_from = 4.0;
38			public static double cutoff_to = 15.0;
39			public static double cutoff_step = 1.0;
40			public static int limit = 100;
41	stehr	250
42	stehr	303	public static String outFileName = "grid_nbs_pdbreps100";
43
44	stehr	250	/---------------------------- private methods --------------------------/
45			/**
46			* Get user name from operating system (for use as database username).
47			* */
48			private static String getUserName() {
49			String user = null;
50			user = System.getProperty("user.name");
51			if(user == null) {
52			System.err.println("Could not get user name from operating system.");
53			}
54			return user;
55			}
56
57			private static void calcDensity(String pdbCode, String chainCode, double cutoff, String egdeType, MySQLConnection conn, Map<Integer, Integer> densityCount) {
58			Pdb pdb = null;
59			try {
60			pdb = new PdbasePdb(pdbCode, chainCode, PDB_DB, conn);
61			// add to density count vector
62			pdb.calcGridDensity(edgeType, cutoff, densityCount);
63
64			} catch (PdbaseInconsistencyError e) {
65			System.out.println("Inconsistency in " + pdbCode + chainCode);
66			} catch (PdbCodeNotFoundError e) {
67			e.printStackTrace();
68			} catch (SQLException e) {
69			e.printStackTrace();
70			} catch (PdbChainCodeNotFoundError e) {
71			e.printStackTrace();
72			}
73
74			}
75
76	stehr	303	public static void printValues(Map<Integer, Integer> v, PrintStream out) {
77	stehr	250	int atoms = 0;
78			for(int size:v.keySet()) {
79	stehr	303	out.println(size + "\t" + v.get(size));
80	stehr	250	atoms += size*v.get(size);
81			}
82	stehr	303	//out.println("Atoms: " + atoms);
83	stehr	250	}
84
85	stehr	303
86			public static void writeResultToFile(Map<Integer, Integer> v, String baseName, String edgeType, double cutoff) {
87			try {
88			File outFile = new File(baseName + "_" + edgeType + "_" + cutoff + ".out");
89			if(outFile.exists()) {
90			outFile.delete();
91			}
92			outFile.createNewFile();
93			printValues(v, new PrintStream(outFile));
94			System.out.println("Results written to file " + outFile.getName());
95
96			} catch (FileNotFoundException e) {
97			e.printStackTrace();
98			} catch (IOException e) {
99			e.printStackTrace();
100			}
101			}
102
103
104			public static void writeResultToDb(Map<Integer, Integer> v) {
105			// insert into runs(run_id, edgeType, cutoff, timestamp, proteins, points, cells, avg_pts_per_cell, max_pts_per_cell, avg_pts_per_area, max_pts_per_area)
106			// insert into density_distr(run_id, points_per_cell, num_cells)
107			// do another run for values per protein (run_id, pdb_id, chain_id, num_res, num_atoms, num_cells, surface_cells, core_cells, surface_area, volume)
108
109			}
110
111			public static Map<Integer, Integer> calcDensity(MySQLConnection conn, String edgeType, double cutoff, boolean verbose) {
112			Map<Integer, Integer> densityCount = new TreeMap<Integer,Integer>();
113	stehr	250	String pdbCode, chainCode;
114			int numPdbs = 0;
115
116	stehr	303	if(verbose) {
117			System.out.print(edgeType + " " + cutoff + ": ");
118			}
119	stehr	250
120			// read structures from database
121	stehr	303	String query = "SELECT DISTINCT " + DB_COL_PDB + "," + DB_COL_CHAIN + " FROM " + DB_NAME + "." + DB_TABLE + " LIMIT " + limit + ";" ;
122	stehr	250	Statement stmt;
123			try {
124			stmt = conn.createStatement();
125			ResultSet rs = stmt.executeQuery(query);
126			while(rs.next()) {
127			pdbCode = rs.getString(1);
128			chainCode = rs.getString(2);
129
130			if(chainCode == null) {
131			chainCode = "NULL";
132			}
133	stehr	303	numPdbs++;
134	stehr	250
135			// calculate statistics
136			calcDensity(pdbCode, chainCode, cutoff, edgeType, conn, densityCount); // will add to densityCount
137	stehr	303	if(verbose) {
138			if(numPdbs %2 == 0) {
139			System.out.print('\b');
140			} else {
141			System.out.print(".");
142			}
143			if(numPdbs % 500 == 0) System.out.print(numPdbs + " ");
144			}
145	stehr	250
146			// for each protein write to db: pdb, chain, num_res, volume, max_density
147			}
148	stehr	303	rs.close();
149			stmt.close();
150			if(verbose) System.out.println(".");
151	stehr	250
152			} catch (SQLException e) {
153			e.printStackTrace();
154	stehr	303	}
155			return densityCount;
156			}
157
158			public static void main(String[] args) {
159			MySQLConnection conn = null;
160			Map<Integer, Integer> densityCount = null;
161			double cutoff;
162
163			// opening db connection
164			try{
165			conn = new MySQLConnection(DB_HOST, DB_USER, DB_PWD);
166			} catch (Exception e) {
167			System.err.println("Error opening database connection. Exiting");
168			System.exit(1);
169	stehr	250	}
170
171	stehr	303	for(cutoff = cutoff_from; cutoff <= cutoff_to; cutoff += cutoff_step) {
172			// run calculation
173			densityCount = calcDensity(conn, edgeType, cutoff, true);
174
175			// output results
176			//printValues(densityCount, System.out);
177			writeResultToFile(densityCount, outFileName, edgeType, cutoff);
178			}
179			System.out.println("Done.");
180
181	stehr	250	}
182
183			}