7.11.2008

来源：互联网发布：effective java mobi 编辑：程序博客网时间：2024/05/16 02:34

1、重装PROFphd

1）运行 ./PROFphd.run

2）使用默认安装

3）修改/prof/bin/ 下的文件后缀名

！！！把 *.LINUX该会 *.UNKNOWN

4）运行测试安装 /prof/scr/test-install.pl

5）预测：

./prof ./exa/1ppt.hssp fileRdb=1ppt.out sec

预测二级结构

结果：

********************************************************************

# Perl-RDB

# PROFsec

# Copyright : Burkhard Rost, CUBIC NYC / LION Heidelberg

# Email : rost@columbia.edu

# WWW : http://cubic.bioc.columbia.edu

# Version : 2000.02

# --------------------------------------------------------------------------------

# About your protein :

# VALUE PROT_ID : 1ppt

# VALUE PROT_NAME : PANCREATIC HORMONE

# VALUE PROT_NCHN : 1

# VALUE PROT_NRES : 36

# VALUE PROT_NALI : 30

# VALUE PROT_NFAR : 25

# VALUE PROT_NFAR50-5: 7

# VALUE PROT_NFAR40-5: 5

# VALUE PROT_NFAR30-5: 4

# VALUE PROT_NFAR5-5: 0

# --------------------------------------------------------------------------------

# About the alignment:

# VALUE ALI_ORIG : ./exa/1ppt.hssp

# --------------------------------------------------------------------------------

# About PROF specifics:

# VALUE PROF_FPAR : sec=/picb/home40/cyd/dys/backup/PROFphd/prof-tmp/prof/net/PROFsec_best.par

# VALUE PROF_NNET : sec=2

# --------------------------------------------------------------------------------

# Notation used :

# ------------------------------------------------------------------------

# NOTATION HEADER : PROTEIN

# NOTATION PROT_ID : identifier of protein [w]

# NOTATION PROT_NAME : name of protein [w]

# NOTATION PROT_NRES : number of residues [d]

# NOTATION PROT_NCHN : number of chains (if PDB protein) [d]

# NOTATION PROT_NALI : number of proteins aligned in family [d]

# NOTATION PROT_NFAR : number of distant relatives [d]

# ------------------------------------------------------------------------

# NOTATION HEADER : ALIGNMENT

# NOTATION HEADER : ALIGNMENT: input file

# ------------------------------------------------------------------------

# NOTATION HEADER : INTERNAL

# NOTATION PROF_FPAR : name of parameter file, used [w]

# NOTATION PROF_NNET : number of networks used for prediction [d]

# ------------------------------------------------------------------------

# NOTATION BODY : PROTEIN

# NOTATION NO : counting residues [d]

# NOTATION AA : amino acid one letter code [A-Z!a-z]

# NOTATION CHN : protein chain [A-Z!a-z]

# ------------------------------------------------------------------------

# NOTATION BODY : PROFsec

# NOTATION OHEL : observed secondary structure: H=helix, E=extended (sheet), blank=other (loop)

# NOTATION PHEL : PROF predicted secondary structure: H=helix, E=extended (sheet), blank=other (loop) PROF = PROF: Profile network prediction HeiDelberg

# NOTATION RI_S : reliability index for PROFsec prediction (0=lo 9=high) Note: for the brief presentation strong predictions marked by '*'

# NOTATION pH : 'probability' for assigning helix (1=high, 0=low)

# NOTATION pE : 'probability' for assigning strand (1=high, 0=low)

# NOTATION pL : 'probability' for assigning neither helix, nor strand (1=high, 0=low)

# NOTATION OtH : actual neural network output from PROFsec for helix unit

# NOTATION OtE : actual neural network output from PROFsec for strand unit

# NOTATION OtL : actual neural network output from PROFsec for 'no-regular' unit

# --------------------------------------------------------------------------------

No AA OHEL PHEL RI_S pH pE pL OtH OtE OtL

1 G L L 9 0 0 9 0 1 97

2 P L L 9 0 0 9 1 2 93

3 S L L 8 0 0 9 2 4 91

4 Q L L 8 0 0 9 1 6 88

5 P L L 8 0 0 9 2 5 91

6 T L L 8 0 0 8 1 9 89

7 Y L L 8 0 0 9 0 10 91

8 P L L 8 0 0 9 2 5 90

9 G L L 8 0 0 8 4 6 87

10 D L L 7 0 0 8 4 7 84

11 D L L 8 0 0 9 1 6 90

12 A L L 9 0 0 9 1 4 95

13 P L L 8 0 0 9 7 1 90

14 V H H 7 9 0 0 87 0 8

15 E H H 8 9 0 0 92 0 4

16 D H H 8 9 0 0 93 0 5

17 L H H 9 9 0 0 93 0 3

18 I H H 8 9 0 0 90 0 4

19 R H H 8 9 0 0 92 0 4

20 F H H 9 9 0 0 94 0 3

21 Y H H 9 9 0 0 95 0 2

22 D H H 9 9 0 0 94 0 3

23 N H H 9 9 0 0 94 0 3

24 L H H 8 9 0 0 93 0 5

25 Q H H 8 9 0 0 92 0 4

26 Q H H 8 9 0 0 89 0 7

27 Y H H 8 9 0 0 91 1 5

28 L H H 8 9 0 0 88 2 7

29 N H H 7 8 0 1 86 2 10

30 V H H 7 8 0 1 85 3 10

31 V H H 3 6 0 2 64 4 27

32 T L L 0 4 0 4 41 6 44

33 R L L 2 3 0 6 27 7 55

34 H L L 3 2 1 6 19 18 56

35 R L L 5 1 1 6 12 18 68

36 Y L L 9 0 0 9 1 3 94

Good!!!!!!!!!!!!!!!!!!!!!!!!!

注意：

PROFphd的input文件可以是fasta，但是格式必须是：

>seq_name

ASJOOEWFNFNSK…………..

多序列预测报错：

5..--> ERROR during initialising copf

msg=*** ERROR copf:brIniErr: after lib-ut:brIniErr

*** ERROR file (fileMatGcg) '/picb/home40/cyd/dys/backup/PROFphd/prof-tmp/phd/mat/Maxhom_McLachlan.metric' missing!

*** ERROR prof:protRd: after protRdAli, fileIn=mutli_seq.fasta,

*** message from sbr:

*** ERROR protRdAli: failed conversion to HSSP

ok lib-col:sysRunProg

*** ERROR prof: line number of error=106

*** ERROR failed prof:full:doOne dbfile=mutli_seq.fasta, chain=*!

*** ERROR msg from where it failed:

*** ERROR doOne: reading db=mutli_seq.fasta, chain=*! (&protRd)

line=1431

*** ERROR protRd: *** ERROR prof:protRd: after protRdAli, fileIn=mutli_seq.fasta,

*** message from sbr:

*** ERROR protRdAli: failed conversion to HSSP

ok lib-col:sysRunProg

新思路：重复调用预测预测单条序列的命令

1、制作input文件，将单条序列名和序列放入一个文件，做成16632个子文件

1）做只有name和seq的文件：

combine.cpp

By DYS

Function:Combine name real structure and prediction structure

Inputfile:name.txt real_structure prediction_structure

Onputfile:result.txt

#include <stdio.h>

#include <fstream>

#include <iostream>

using namespace std;

int main()

{

ifstream name("name.txt");

ifstream aa("protein.txt");

ofstream combine("name_protein.txt");

string name_str,aa_str;

name>>name_str;

while(!name.eof())

{

aa>>aa_str;

combine<<name_str<<endl<<aa_str<<endl;

name>>name_str;

}

}//int main

注意：expr的用法，expr num1 + num2 两数与加号之间要空开