FOR C++ EXPERTS - need your help (Jan/06/2007 )
I think if you are just starting programming then you should look into perl. The way you wrote the program initially is valid, and has the syntax of perl/c. Perl is easy to learn, and you shouldn't have a problem with it if you already know some c++. Trying to wrap your head around the object-orientated c++, which is what perlmunky is showing how to write, is not the best starting point.
hi jaknight...i told u previously that i'm just a beginner in programming, i chose to work on C++ because i have a small background about it from one course i studied years ago..so i need more practice to refresh my memory and become used to it, and since perl is not widely different from C++ as u said, i will start learning it as soon as i can...
I actually totally disagree with you (jaknight). The best place to start is with OO and a language like C++/JAVA. This is because it is strict, you can't get away with hacks (as you can with perl). The rule set is much better defined. I started out with python using it in an semi-object orientated way (the way python is written it is alway OO) I didn't create my own objects. I then did perl before C/C++/JAVA - I found the transition far harder than my friends who had experience in OO languages.
However, this task is a perl/scripting language job - 'berry however has some experience of c/c++ and wants to continue down that route so I don't see the point of either of us trying to turn 'berry to the light side.
If you can get the basics of c/c++ you will find perl a walk in the park. IMHO
For an easy-to-learn, strongly-typed, OOP language, how about Ruby?
My preference would be Perl, because I know it, and it seems to be a very good choice for bioinformatics -- there are a lot of modules around geared toward bio, and many people in biology use it. Also, since Per is a scripted language (as is Ruby), there are no compiler issues to deal with, so it's easier for cross-platform compatability.
But, I also agree with perlmunky to some extent -- learn what you're most comfortable with...
oops... I forgot ruby (i am trying to find time to learn it). From the little I have used it, it seems very nice. Not too far from standard english (unlike C, but like java).
see: RUBY - from a Digg link. There is also ruby on rails for web development and so on.
i modified the program, using one array only by calling a function applying switch:
#include<iostream>
using namespace std;
void rna(char seq[], int size)
{
for(int j=0;j<=size;j++)
{
switch(seq[j]){
case 'A':
seq[j]='U';
break;
case 'T':
seq[j]='A';
break;
case 'G':
seq[j]='C';
break;
case 'C':
seq[j]='G';
break;
default :
cout<<" "<<endl;
break;
}
}
}
int main()
{
char s1[20];
int t,n,mp;
int i, count1=0,count2=0,count3=0,count4=0;
cout<<"enter your dna sequence: \n";
cin>>s1;
for (i=0;s1[i]!='\';i++)
{
if (s1[i]=='A')
count1=count1+1;
else if (s1[i]=='T')
count2=count2+1;
else if(s1[i]=='G')
count3=count3+1;
else
count4=count4+1;
}
n=count1+count2+count3+count4;
cout<<"total length of the sequence= "<<n<<endl;
cout<<"number of amino acids= "<<n/3<<endl;
cout<<"%A= "<<count1*100/n<<endl;
cout<<"%T= "<<count2*100/n<<endl;
cout<<"%G= "<<count3*100/n<<endl;
cout<<"%C= "<<count4*100/n<<endl;
t=(count3+count4)*100/n;
cout<<"GC%= "<<t<<endl;
mp=2*(count1+count2)+4*(count3+count4);
cout<<"melting point of this sequence= "<<mp<<endl;
rna( s1, n);
cout<<s1<<endl;
return 0;
}
Hi,
that looks good, however you still have another pointless loop.
{
if (s1[i]=='A')
count1=count1+1;
else if (s1[i]=='T')
count2=count2+1;
else if(s1[i]=='G')
count3=count3+1;
else
count4=count4+1;
Rather than use a function you should build an object. Remember and object contains a group of functions.
Because of the scale of this bit of work you are probably best using a single c style program that could later be used as an object.
Rather than having individual counters (count 1 count 2) and so on, create an array to hold counters or better yet a structure.
where:
int [] counters = new int [4]
where counters[0] => the number of Ts
counters[1] => the number of As and so on.
Again in your switch case setup, increment the correct array position
case 'A':
seq[j]='U';
counters[1]++;
break;
case 'T':
seq[j]='A';
counters[0]++;
break;
This will save you more time in more complex tasks.
Also, to make you code easier to read, I find that using proto-types is good. This means that you define your function at the top of the program, but build the major code after the main(){ }
void dumpCharArray(char array []);
int main {
char [] array = new char [40];
}
##########################
# #
# FUNCTIONS #
# #
##########################
void dumpCharArray(char array []) {
for (int i = 0; i < array.size; i++) { //size might be a python or java method not sure//
std::cout << "Current base is " << array[i] << endl;
}
}
Something like that.
THANX perlmunky for your consideration
i am still a beginner , still have to learn objects , structures and files and that's why i use simple design
be patient with me 
using namespace std;
int main()
{
char seq[20];
int size=20;
int t,n,mp;
int counta=0,countt=0,countg=0,countc=0;
cout<<"enter your dna sequence: \n";
cin>>seq;
for(int j=0;j<=size;j++)
{
switch(seq[j]){
case 'A':
seq[j]='U';
counta++;
break;
case 'T':
seq[j]='A';
countt++;
break;
case 'G':
seq[j]='C';
countg++;
break;
case 'C':
seq[j]='G';
countc++;
break;
default :
cout<<" "<<endl;
break;
}
}
n=counta+countt+countg+countc;
cout<<"total length of the sequence= "<<n<<endl;
cout<<"number of amino acids= "<<n/3<<endl;
cout<<"%A= "<<counta*100/n<<endl;
cout<<"%T= "<<countt*100/n<<endl;
cout<<"%G= "<<countg*100/n<<endl;
cout<<"%C= "<<countc*100/n<<endl;
t=(countg+countc)*100/n;
cout<<"GC%= "<<t<<endl;
mp=2*(counta+countt)+4*(countg+countc);
cout<<"melting point of this sequence= "<<mp<<endl;
cout<<"mRNA sequence is "<<seq;
return 0;
}
That's exactly what I meant.
You could maybe create an array instead of the individual counts. Then rather than having four individual print statments (to show % contribution of each acid) you could have a function that takes an array and for each element calculates the percentage contribution of said element.
Don't forget to annotate your code. I find it helps. If you do get around to using perl have a look at perldoc, I am not sure if there is a C/C++ version.
I am happy to help where I can: basically perl, c/c++/ , java, python, bash and ruby (soon I hope)
oh also, get yourself a copy of eclipse and the language plug-ins it makes life organised and writing far easier.
here is my modified small program 
using namespace std;
void mycalculation(int counter[], int size)
{
int n,mpoint,sum=0;
float i;
for(int j=0;j<size;j++){
sum=sum+counter[j];}
cout<<"\nthis sequence has a length of "<<sum<<endl;
cout<<"\nthe % of A,T,G,C ARESPECTIVELY is"<<endl;
for(j=0;j<size;j++){
n=counter[j]*100/sum;
cout<<n<<"%"<<endl;}
i=(counter[2]+counter[3])*100/sum;
cout<<"\nCG%= "<<i<<endl;
mpoint=(counter[0]+counter[1])*2+(counter[2]+counter[3])*4;
cout<<"\nmelting point for this sequence= "<<mpoint<<endl;
cout<<"\nexpected number of amino acids translated= "<<sum/3<<endl;
}
int main()
{
char sdnaseq[100];char sdnaseq2[100];char rnaseq[100];char doubls[2][100];
int i,base=4;int size=100;
int countbase[4]={0};
cout<<"this is a single stranded DNA sequence:"<<endl;
cin>>sdnaseq;
for(i=0;i<size;i++)
{
switch(sdnaseq[i]){
case 'A':
sdnaseq2[i]='T';
rnaseq[i]='A';
countbase[0]++;
break;
case 'T':
sdnaseq2[i]='A';
rnaseq[i]='U';
countbase[1]++;
break;
case 'G':
sdnaseq2[i]='C';
rnaseq[i]='G';
countbase[2]++;
break;
case 'C':
sdnaseq2[i]='G';
rnaseq[i]='C';
countbase[3]++;
break;
default:
sdnaseq2[i]=0;
rnaseq[i]=0;
break;
}
}
for(i=0;i<100;i++){
doubls[0][i]=sdnaseq[i];
doubls[1][i]=sdnaseq2[i];}
cout<<"\ndouble stranded DNA: "<<endl;
for(i=0;i<2;i++){
for(int j=0;j<100;j++){
cout<<doubls[i][j];}
}
cout<<"\nif the coding DNA strand is: "<<sdnaseq2<<" \ntherefore the mRNA transcript would be: "<<rnaseq<<endl;
mycalculation(countbase,base);
return 0;
}