This is an R package that calculates alignment-free vectors, in particular the generalized vector, for dna sequences. Additionally, there are functions to extract relevent information from fasta and gbk files and to compute weighted skewness and kurtosis.
An existing Bioconductor package, Biostrings, can extract kmer counts, but leaves you to calculate the frequencies or other statistics. Biostrings also ignores ambiguous nucleotides such as R, B, or N. Additionally, there are no existing functions to extract sequence information from gbk files.
The functions gbk and fasta extract sequence and other information from gbk and fasta files, respectively. The Vectorizer function will convert the DNA or RNA character string into the generalized vector described here: http://arxiv.org/abs/1309.0408
The proper configuration of the options can allow the Vectorizer funtion to compute kmer frequencies only, the natural vector, and an additional function to convert kmers to the composition vector.