Author

Brent Pedersen (brentp)

Email

bpederse@gmail.com

License

MIT

4bit encoding of DNA sequence with alphabet of: "-ACGNTXacgntx" other characters are encoded as 'N'. The encoding works by putting 2 chars into a single unsigned char. Since there are 255 unsigned chars, and only 12 * 12 = 144 pairs in our alphabet, this works simply. The code is contained entirely in the file: '4bit.h' with the API function signatures defined and documented below.

encode a dna sequence string into a string of unsigned chars. the returned value must be free'd. :

const unsigned char *encode(unsigned char *dna_seq);

encode the given dna_seq string and write it to the file. the return value is the end fseek position of the written, encoded sequence :

size_t fencode(FILE *fh, unsigned char *seq);

decode an encoded value into it's dna sequence string. the return value must be free`d :

const unsigned char* decode(unsigned char *positions);

return the dna sequence string encoded in the file at (0-based) file positions fh_start to fh_end inclusive. the return value must be free'd. :

const char* fdecode(FILE *fh, size_t fh_start, size_t fh_end);

typedef struct f4bit {
    size_t bp_start; // 0 -based
    size_t bp_end; // 0 -based
    size_t fh_start;
    size_t fh_end;
} f4bit;

identical to fencode except the return value is an f4bit struct pointer containing the base-pair and file-handle start and stop of the data. useful for creating an index to save where particular chromosomes start and stop within the file. the return value must be free`d. :

const f4bit* fencodes(FILE *fh, unsigned char *seq);

in progress. see 4bit-cli.c and test.sh

Please report any bugs, suggestions or code improvements. For example use and tests, see the test-4bits.c file.