written by [email protected]
 
 (1) takes in a bed file (can have any number of columns).
         - optionally finds midpoint and extends features (requires chr_sizes file found from species string)
         - optionally just extends each feature (requires chr_sizes file found from species string)
         
 (2) extract sequence data as a fasta file (requires path to whole_genome_fasta_file; entered as a string)
     fasta file has no text wrapping (i.e. entire sequence for an id is on one line)
 
 (3) for each each feature in the fasta file it:
         - counts number of observed CpG dinucleotides
         - calculates the expected number of CpG (#C * #G)/length of sequence
         - calculates CpG observed/expected
         
 (4) inserts the CpG o/e into the original supplied bedfile and returns this as a pandas dataframe. 
 
 Note that the calculated CpG o/e is for the genomic region specified by the supplied parameters (e.g. extending flanks),
 but the original bed file co-ordinates are returned