![]() |
|
![]() |
||||
Thread | Thread Starter | Forum | Replies | Last Post |
Software tools for detecting Copy Number Variations (CNV) | smice | Bioinformatics | 26 | 05-04-2016 09:05 AM |
CNV-seq, to detect Copy Number Variation using next-generation sequencing | xiechao | Literature Watch | 46 | 10-20-2015 09:30 AM |
Looking for the right WGS simulator | oiiio | Bioinformatics | 5 | 07-20-2012 10:59 AM |
PubMed: ART: a next-generation sequencing read simulator. | Newsbot! | Literature Watch | 0 | 12-27-2011 02:00 AM |
Nexgen simulator. | aloliveira | General | 1 | 02-14-2011 08:57 AM |
![]() |
|
Thread Tools |
![]() |
#1 |
Member
Location: India Join Date: Aug 2010
Posts: 78
|
![]()
Hello,
I have written a program to simulate CNV, into the reference genome. It take the Reference Genome/Chromosome as input. The program then generates a Random number(X), copies the segment from that position(X), size specified by user. Then generates another random number(Y), and inserts the copied segment from X position, at Y position. The output is the modified input file with inserted CNV and a txt file with Location of CNVs ie. position where copied from (X) and where Inserted (Y). I give the Random Number generation function MIN and MAX value. MIN =1 and MAX=last Position(Size), it then generates a Random number within these limits. Problem is, with same input(MIN & MAX) the random numbers generated is same. eg. If I make my program run once and generate 100 numbers, they are random. But if I generate 100 numbers by running the program again, the numbers are same as before. CODE: //call this function first to seed void initialize_rand(){ srand((unsigned)(time(0))); } //generates a psuedo-random integer between min and max long int randint(long int min, long int max){ if(min>max){ return lrand48()%(min-max)+max; } else{ return lrand48()%(max-min)+min; } } |
![]() |
![]() |
![]() |
#2 |
Member
Location: India Join Date: Aug 2010
Posts: 78
|
![]()
I know this is not a C programming forum.
Googled the problem, but didn't find anything useful. But as many of the member here have dealt with similar situation, hence I thought of asking here. |
![]() |
![]() |
![]() |
#3 |
Simon Andrews
Location: Babraham Inst, Cambridge, UK Join Date: May 2009
Posts: 871
|
![]()
If that's happening then it means your RNG is using the same seed value each time. I'm not a C programmer but I suspect that time(0) returns the same value each time and that you probably want to use time(NULL) to get the current time value so the seed is different each time.
|
![]() |
![]() |
![]() |
#4 |
Senior Member
Location: Charlottesville Join Date: Sep 2008
Posts: 119
|
![]()
Your issue is likely that the time() function has granularity to the second. If you make multiple calls to you program in the same second, the results will be identical. The following will seed based on time and process_id on POSIX systems. Therefore, for a given boot cycle, you are guaranteed not to have redundant seeds
Code:
srand((unsigned)time(0)+(unsigned)getpid()) Code:
#include <unistd.h> Last edited by quinlana; 01-28-2011 at 10:20 AM. Reason: typo |
![]() |
![]() |
![]() |
#5 |
Member
Location: India Join Date: Aug 2010
Posts: 78
|
![]()
Thank you everyone for the reply.
I made the change suggested by "quinlana". Also changed the seeding function to srand48(). Every thing is working fine now. |
![]() |
![]() |
![]() |
#6 |
Member
Location: singapore Join Date: Apr 2012
Posts: 12
|
![]()
is your program available for download? i would like to generate simulated CNV
|
![]() |
![]() |
![]() |
Thread Tools | |
|
|