1. Stand-alone tools 2. 1.Download the zip file to the GMS6014 folder. 2.Unzip the files to a folder named “clustalx”. 3.Edit the MDM2_isoforms_5.fasta file with WordPad and save. 4.Run the.exe file. 5.Load sequence file, select sequences, perform alignment. 6.Write the alignment to a ps file. Practice –the ClustalX application.

2. Stand-alone tools 3. Command line applications:  Accounts for a large number of high-quality, sophisticated programs. Practice – (install and) run standalone blast in your own computer

3. Searching for potential ortholog of oncogene MDM2 in the fruit fly genome Pet Projects:

4. Practice – Install the blast program (1) 1.Download the BLAST executable file, save the file in a folder, such as c:\GMS6014\blast\ 2.Run the installation program by double click. Inspect the folder following installation. 3.Add three more folders to your /blast directory, “/query”, “/dbs”, and “/out”.

5. Practice – Install the blast program (2) 5.Inspect the contents of the doc, data, and bin folder. Move the programs from blast\bin to the blast folder. 6.Bring a command (cmd) window by typing “cmd” in the Start  Run box. 7.Go to the blast folder by typing “cd C:\GMS6014\blast” 8.Try to run the program by typing “blastall”, read the output.

6. Practice -- BLAST search in your own computer 1.Download data file from the course web page, or Ensemble. Save in the blast\dbs folder. 2.Start a CMD window, navigate to the C:\GMS6014\blast folder. 3.At the prompt “C:\GMS6014\blast >” type the command “formatdb –i dbs\Dm.P –p T” -- format the dataset for the program. 4.Compose the query sequence save as “3TNF.txt” in the “blast\query\” folder. 5.Initiated the search by typing “blastall –p blastp –d dbs\Dm.P –i query\4_MMD2.fasta –o out\Mdm2_DmP.html –T T”

7. What’s in a command? formatdb –i dbs\Dm.P –p T Program – format database for search. Feed me the input file name Tell me is it a protein sequence file? For more info, refer to the “user manual” file in the blast\doc folder.

8. Advantages of Running BLAST at Your Own Machine  Do it at any time, no waiting on the line.  Search for multiple sequences at once.  Search a defined data set.  Automate Blast analysis.  Combine Blast with other analysis.  …..

9. BLAST is a program implemented in C/C++ void BlastTickProc(Int4 sequence_number, BlastThrInfoPtr thr_info) { if(thr_info->tick_callback && (sequence_number > (thr_info->last_db_seq + thr_info->db_incr))) { NlmMutexLockEx(&thr_info->callback_mutex); thr_info->last_db_seq += thr_info->db_incr; thr_info->tick_callback(sequence_number, thr_info->number_of_pos_hits); thr_info->last_tick = Nlm_GetSecs(); NlmMutexUnlock(thr_info->callback_mutex); } return; } /* Sends out a message every PERIOD (i.e., 60 secs.) for the index. THis function runs as a separate thread and only runs on a threaded platform. Should I care ?

10. If you care: 1.) Data structure and Algorithm char: name char: sequence SEQ Identify the best alignment for two sequences (p69-73) Seq1: MA-DSV—WC.. Seq2: MALD-IHWS.. int: seq_length

11. Programming language comparison /* TRANSLATION: 3 or 6 frame translate cDNA sequences */ //--------------------------------------------------------------------------- #include "translation.hpp" int main(int argc, char **argv) { int num_seq=0; char string[MAXLINE]; DSEQ * dseq; infile.getline (string,MAXLINE); if (string[0]=='>') strncpy (dbname,string,MAXLINE); while (!infile.eof()) { dseq=Get_Lib_Seq (); if (dseq->reverse==0) Translation (&dseq->name[1], dseq->seq); else Translation (&dseq->name[1], dseq->r_seq); num_seq++; if (num_seq%1000==0) { cout<<num_seq<<endl; cout name<<endl; } delete dseq; } infile.close(); outfile.close(); cout<<num_seq<<" translated"<<endl; getch(); return 0; } DSEQ* Get_Lib_Seq() { int i,n; char str[MAXLINE]; DSEQ* dseq; n = 0; dseq=new DSEQ; strcpy (dseq->name, dbname); while(infile.getline(str,MAXLINE)) {if (str[0] == '>') { strcpy( dbname, str); break; } for(i=0;i<strlen(str);i++) {if(n==MAXSEQ) break; dseq->seq[n++] = str[i]; } dseq->seq[n]='\0'; if(n==MAXSEQ) cout<<"WARNING: sequence"<<dbname<<"too long!"<<endl; dseq->len=n; if (dseq->name[9]=='3') Reverse (dseq); else dseq->reverse=0; return dseq; } void Reverse (DSEQ* dseq) //Reverse dseq {int i,j; j=0; for (i=(dseq->len-1);i>0;i--) { if (dseq->seq[i]=='A'||dseq->seq[i]=='a') dseq->r_seq[j++]='T'; if (dseq->seq[i]=='C'||dseq->seq[i]=='c') dseq->r_seq[j++]='G'; if (dseq->seq[i]=='G'||dseq->seq[i]=='g') dseq->r_seq[j++]='C'; if (dseq->seq[i]=='T'||dseq->seq[i]=='t') dseq->r_seq[j++]='A'; if (dseq->seq[i]=='N'||dseq->seq[i]=='n') dseq->r_seq[j++]='N'; } dseq->r_seq[j++]='\0'; dseq->reverse=1; } void Translation (char name[], char seq[]) { char ppseq[MAXSEQ/3]; for (int f=0; f<3; f++) { outfile "<<"F_"<<f<<name<<endl; int j=0; int len=strlen(seq); for( int i=f; i<len; i=i+3) ppseq[j++]=Translate(&seq[i]); ppseq[j++]='\0'; int m=strlen(ppseq)/50; // output 50 aa per line for (int n=0; n<=m; n++) {for (int i=n*50; i<50*(n+1); i++) {outfile<<ppseq[i]; if (ppseq[i]=='\0') break; } outfile<<endl; } char Translate(char s[]) { int c1,c2,c3; char P, code[3]; //***standard translation table, A(0),C(1), G(2), T(3)***** char table [4][4][4]= {{{'K','N','K','N'},{'T','T','T','T'},{'R','S','R','S'},{'I','I','M','I'}}, {{'Q','H','Q','H'},{'P','P','P','P'},{'R','R','R','R'},{'L','L','L','L'}}, {{'E','D','E','D'},{'A','A','A','A'},{'G','G','G','G'},{'V','V','V','V'}}, {{'*','Y','*','Y'},{'S','S','S','S'},{'*','C','W','C'},{'L','F','L','F'}}}; //*********** table2 for n at 3rd position******************** char table2 [4][4]={{'X','T','X','X'},{'X','P','R','L'}, {'X','A','G','V'},{'X','S','X','X'}}; strncpy (code, s, 3); c1=Convert(code[0]); c2=Convert(code[1]); c3=Convert(code[2]); if (c1>=4 || c2>=4) P='X'; //can be Optimized further here by considering.... else { if (c3>=4) P=table2[c1][c2]; else P=table[c1][c2][c3]; //P=table[Convert(code[0])][Convert(code[1])][Convert(code[2 ])]; } return (P); } int Convert (char c) { char s=c; if (s=='A'||s=='a') return (0); if (s=='C'||s=='c') return (1); if (s=='G'||s=='g') return (2); if (s=='T'||s=='t'||s=='U'||s=='u') return (3); if (s=='N'||s=='n') return (4); else return (5); } f#Translation -- read from fasta DNA file and translate into three frames # import string from Bio import Fasta from Bio.Tools import Translate from Bio.Alphabet import IUPAC from Bio.Seq import Seq ifile = "S:\\Seq\\test.fasta" parser = Fasta.RecordParser() file =open (ifile) iterator = Fasta.Iterator (file, parser) cur_rec = iterator.next() cur_seq = Seq (cur_rec.sequence,IUPACUnambiguousDNA()) translator = Translate.unambiguous_dna_by_id[1] translator.translate (cur_seq) Translation : C Translation : Python

12. Programming languages C/C++ Java - Biojava Python - Biopython Perl - Bioperl Efficiency, Power Simplicity, Fast Dev.

13. Observe: scripting is not that difficult Example: Python and bioPython. 1.Simple python scripts. 2.Batch Blast with a Python script.

14. Blast outputoutput

15. Questions after the Blast search? Questions: Is this a expressed gene in the Fruit fly? - Gene prediction & gene structure Is this the true ortholog of MDM2? - Fundamentals of sequence comparison What can we learn from the comparison of sequences? -- protein dommains/motifs.

16. Blast outputoutput

17. How to measure the similarity between two sequences Q: which one is a better match to the query ? Query: M A T W L Seq_A: M A T P P Seq_B: M P P W I

18. Judging the match using “Scoring Matrix” Q: which one is a better match to the query ? Query: M A T W L Seq_B: M P P W I Score: 5-1-1 112 Total: 16 Total: 7 -4 Query: M A T W L Seq_A: M A T P P Score: 545-3

19. “Scoring Matrix” assigns a score to each pair of amino acids A S T L I V K D... L –1 –2 –2 4 3 1 -2 –4 BLOSUM-62

20. BLOSUM - Blocks Substitution Matrices Block: very well conserved region of a protein family. – perform the same (similar) function. ASLDEFL SALEDFL ASLDDYL ASIDEFY ASIDEFY … Score(a1/a2) = observed frequency of a1/a2 2* log2 predicated frequency of a1/a2 AA: 6 AS: 3 SS: 0

21. BLOSUM - Blocks Substitution Matrices Block: very well conserved region of a protein family. – perform the same (similar) function. ASLDEFL ASLEDFL ASLDDYL SALEEFL ASLDDYL SALEEFL … Score (a1/a2) > 0 = 0 < 0 observed frequency of a1/a2 > predicated frequency of a1/a2 observed frequency of a1/a2 predicated frequency of a1/a2 <

22. BLOSUM - Blocks Substitution Matrices Block: very well conserved region of a protein family. – perform the same (similar) function. ASLDEFL ASLEDFL ASLDDYL SALEEFL ASLDDYL SALEEFL … observed frequency of L / I > predicated frequency of L / I i.e: 0.1*0.1 = 0.01i.e: 0.03 Score (L/I) > 0 Substitution of L / I is common in conserved sequences

23. BLOSUM - Blocks Substitution Matrices Block: very well conserved region of a protein family. – perform the same (similar) function. ASLDEFL ASLEDFL ASLDDYL SALEEFL ASLDDYL SALEEFL … observed frequency of L / K < predicated frequency of L / K i.e: 0.1*0.1 = 0.01i.e: 0.0002 Score (L/K) < 0 Substitution of L / K is rare in conserved sequences

24. “Scoring Matrix” assigns a score to each pair of amino acids A S T L I V K D... L –1 –2 –2 4 3 1 -2 –4 BLOSUM-62

25. Scoring matrix –BLOSUM 62