The last problem in the first set of Software Carpentry goes like this:

"9. In genetic cryptography, the four bases ACGT represent the digits 0123 in a base-4 numerical system, so CGA is (1*4^2)+(2*4)+0 = 20. Write a function dna2num(str) to convert a DNA string to a number, and another function num2dna(int) to convert an integer to a DNA string."

There's a little typo there -- the base-4 result 20 should read 120 -- but never mind that. The bigger issue is that you can't convert every DNA string to an integer and back. If your string starts with an A base, as in ACGA, you need a zero-padded string representation of the equivalent integer -- i.e., the string "0120", not the number 120. So I decided that the function num2dna(int) would work better if it took a string argument instead -- num2dna(str).

Also, the desired behavior is left to your discretion. Do you want to be able to convert base-10 numbers to DNA strings, or do you want your function to return an error if, as it parses the input string, it encounters a digit that is larger than or equal to 4? Do you want to assume that an input string with such a digit is base-10, as opposed to some other base greater than 4?

My solution assumed that any input string that was not a zero-padded base-4 integer would be a zero-padded base-10 integer. I wanted the num2dna(str) function to simply make a note of this, convert it to base-4, and take it from there. I also wanted the function dna2num(str) to output a base-10 string.

Below is the set of functions I wrote for this conversion, after going through the first day's worth of lectures, a bit of Google help with base conversion, and a quick stop at Stack Overflow:

# problem 9 dna to number, number to dna
# first, three helper functions: fromb4tob10, fromb10tob4, checkb4

def fromb4tob10(numstr): # convert padded base-4 to padded base-10 string
   for i in range(len(numstr)):
   return "%0*d" % (len(numstr),j)

def fromb10tob4(numstr): # convert padded base-10 to base-4 string
   while (myint-myint%4)>0:
   return "%0*d" % (len(numstr),int(mystr))

def checkb4(dnanumstr): # if input string is base-10, convert to base-4
   for i in range(len(dnanumstr)):
      if int(dnanumstr[i])>=4:
         print "input string is base 10, will change to base 4"
         return fromb10tob4(dnanumstr)
   return dnanumstr

def dna2num(str):
   for i in range(len(str)):
      for base in bases:
         if str[i]==base:
      if check==0: # if this letter is not A,C,G, or T, then exit.
         print "this is not a valid dna sequence"
         return 0
         for base in bases:
            if str[i]==base:
   return fromb4tob10("%0*d" % (len(str),num))

def num2dna(dnanumstr):
   print "input string translates to (padded) base-4 integer: ", "%0*d" % (len(dnanumstr),n)
   s=str(n) # will need this to check if input string is padded with zeroes
   if len(s)!=len(dnanumstr):              # in that case, we should
      dnastart='A'*(len(dnanumstr)-len(s)) # intialize dnastring with
   bases=['A','C','G','T']                 # the same number of A's
   while (n-n%10)>=0:
      for i in range(len(digits)):
         if num==digits[i]:
      if n==0:
         return dnastart+dnastring

I saved the code above as and tested it in SciTE with a few tasks that I thought would reveal any unexpected behavior, as follows:

import functions as f
print f.fromb10tob4('024')
print f.fromb4tob10('120')
print f.fromb4tob10(f.fromb10tob4('024'))
print "must see same numbers on two rows above this"
print f.dna2num(f.num2dna("024"))
print f.dna2num("ACGA")
print f.dna2num("FGA")
print f.dna2num("ACGTTGCAAGCTCGA")
print f.num2dna("4")
print f.num2dna("123")
print f.num2dna("00123")
print f.num2dna("000000116984280")
print f.num2dna(f.dna2num("ACGTTGCAAGCTCGA"))
print f.num2dna(f.dna2num("CCC"))
print f.num2dna(f.dna2num("AAAA"))
print "homework problem 9"

It looks like it works OK. If you're into Python and you would have done any of this differently, I am curious. Meanwhile, I'll plod ahead with the rest of the lectures. I'm liking Software Carpentry.