Kmer.cpp

#include "Kmer.hpp"


// use:  int2bin(a, buffer, buf_size);
// pre:  buf_size >= 8 and buffer has space for buf_size elements
// post: buffer[0,...,7] is the binary representation of a
void int2bin(uint32_t a, char *buffer, int buf_size) {
  //buffer += (buf_size - 1);

  for (int i = 7; i >= 0; i--) {
    *buffer++ = (a & 1) + '0';
    a >>= 1;
  }
}

static const uint8_t base_swap[256] = {
  0x00, 0x40, 0x80, 0xc0, 0x10, 0x50, 0x90, 0xd0,
  0x20, 0x60, 0xa0, 0xe0, 0x30, 0x70, 0xb0, 0xf0,
  0x04, 0x44, 0x84, 0xc4, 0x14, 0x54, 0x94, 0xd4,
  0x24, 0x64, 0xa4, 0xe4, 0x34, 0x74, 0xb4, 0xf4,
  0x08, 0x48, 0x88, 0xc8, 0x18, 0x58, 0x98, 0xd8,
  0x28, 0x68, 0xa8, 0xe8, 0x38, 0x78, 0xb8, 0xf8,
  0x0c, 0x4c, 0x8c, 0xcc, 0x1c, 0x5c, 0x9c, 0xdc,
  0x2c, 0x6c, 0xac, 0xec, 0x3c, 0x7c, 0xbc, 0xfc,
  0x01, 0x41, 0x81, 0xc1, 0x11, 0x51, 0x91, 0xd1,
  0x21, 0x61, 0xa1, 0xe1, 0x31, 0x71, 0xb1, 0xf1,
  0x05, 0x45, 0x85, 0xc5, 0x15, 0x55, 0x95, 0xd5,
  0x25, 0x65, 0xa5, 0xe5, 0x35, 0x75, 0xb5, 0xf5,
  0x09, 0x49, 0x89, 0xc9, 0x19, 0x59, 0x99, 0xd9,
  0x29, 0x69, 0xa9, 0xe9, 0x39, 0x79, 0xb9, 0xf9,
  0x0d, 0x4d, 0x8d, 0xcd, 0x1d, 0x5d, 0x9d, 0xdd,
  0x2d, 0x6d, 0xad, 0xed, 0x3d, 0x7d, 0xbd, 0xfd,
  0x02, 0x42, 0x82, 0xc2, 0x12, 0x52, 0x92, 0xd2,
  0x22, 0x62, 0xa2, 0xe2, 0x32, 0x72, 0xb2, 0xf2,
  0x06, 0x46, 0x86, 0xc6, 0x16, 0x56, 0x96, 0xd6,
  0x26, 0x66, 0xa6, 0xe6, 0x36, 0x76, 0xb6, 0xf6,
  0x0a, 0x4a, 0x8a, 0xca, 0x1a, 0x5a, 0x9a, 0xda,
  0x2a, 0x6a, 0xaa, 0xea, 0x3a, 0x7a, 0xba, 0xfa,
  0x0e, 0x4e, 0x8e, 0xce, 0x1e, 0x5e, 0x9e, 0xde,
  0x2e, 0x6e, 0xae, 0xee, 0x3e, 0x7e, 0xbe, 0xfe,
  0x03, 0x43, 0x83, 0xc3, 0x13, 0x53, 0x93, 0xd3,
  0x23, 0x63, 0xa3, 0xe3, 0x33, 0x73, 0xb3, 0xf3,
  0x07, 0x47, 0x87, 0xc7, 0x17, 0x57, 0x97, 0xd7,
  0x27, 0x67, 0xa7, 0xe7, 0x37, 0x77, 0xb7, 0xf7,
  0x0b, 0x4b, 0x8b, 0xcb, 0x1b, 0x5b, 0x9b, 0xdb,
  0x2b, 0x6b, 0xab, 0xeb, 0x3b, 0x7b, 0xbb, 0xfb,
  0x0f, 0x4f, 0x8f, 0xcf, 0x1f, 0x5f, 0x9f, 0xdf,
  0x2f, 0x6f, 0xaf, 0xef, 0x3f, 0x7f, 0xbf, 0xff
};


// use:  km = Kmer();
// pre:
// post: the DNA string in km is AA....AAA (k times A)
Kmer::Kmer() {
  memset(bytes,0,MAX_K/4);
}


// use:  _km = Kmer(km);
// pre:  s[0],...,s[k] are all equal to 'A','C','G' or 'T'
// post: the DNA string in _km and is the same as in km
Kmer::Kmer(const Kmer& o) {
  memcpy(bytes,o.bytes,MAX_K/4);
}


// use:  km = Kmer(s);
// pre:  s[0],...,s[k] are all equal to 'A','C','G' or 'T'
// post: the DNA string in km is now the same as s
Kmer::Kmer(const char *s) {
  set_kmer(s);
}


// use:  _km = km;
// pre:
// post: the DNA string in _km and is the same as in km
Kmer& Kmer::operator=(const Kmer& o) {
  if (this != &o) {
    memcpy(bytes, o.bytes, MAX_K/4);
  }
  return *this;
}


// use:  km = Kmer();
// pre:
// post: The last bit in the bit array which stores the DNA string has been set to 1
//       which indicates that the km is invalid
void Kmer::set_deleted() {
  memset(bytes,0xff,MAX_K/4);
}


// use:  b = (km1 < km2);
// pre:
// post: b is true <==> the DNA strings in km1 is alphabetically smaller than
//                      the DNA string in km2
bool Kmer::operator<(const Kmer& o) const {
  for (size_t i = 0; i < k_bytes-1; ++i) {
    if (base_swap[bytes[i]] > base_swap[o.bytes[i]]) {
      return false;
    } else if (base_swap[bytes[i]] < base_swap[o.bytes[i]]) {
      return true;
    }
  }

  if (base_swap[bytes[k_bytes-1] & k_modmask] >= base_swap[o.bytes[k_bytes-1] & k_modmask]) {
    return false;
  }
  return true;
}


// use:  b = (km1 == km2);
// pre:
// post: b is true <==> the DNA strings in km1 and km2 are equal
bool Kmer::operator==(const Kmer& o) const {
  return memcmp(bytes,o.bytes,MAX_K/4)==0;
}


// use:  km.set_kmer(s);
// pre:  s[0],...,s[k-1] are all 'A','C','G' or 'T'
// post: The DNA string in km is now equal to s
void Kmer::set_kmer(const char *s)  {
  size_t i,j,l;
  memset(bytes,0,MAX_K/4);

  for (i = 0; i < k; ++i) {
    j = i % 4;
    l = i/4;
    assert(*s != '\0');

    switch(*s) {
    case 'A': break;
    case 'C': bytes[l] |= (0x01 << (2*j)); break;
    case 'G': bytes[l] |= (0x02 << (2*j)); break;
    case 'T': bytes[l] |= (0x03 << (2*j)); break;
    }

    s++;
  }
}


// use:  i = km.hash();
// pre:
// post: i is the hash value of km
uint64_t Kmer::hash() const {
  uint64_t ret;
  MurmurHash3_x64_64((const void *)bytes,k_bytes,0,&ret);
  return ret;
}


// use:  rep = km.rep();
// pre:
// post: rep is km.twin() if the DNA string in km.twin() is alphabetically smaller than
//       the DNA string in km, else rep is km
Kmer Kmer::rep() const {
  Kmer tw = twin();
  return (tw < *this) ? tw : *this;
}


// use:  tw = km.twin();
// pre:
// post: tw is the twin kmer with respect to km,
//       i.e. if the DNA string in km is 'GTCA'
//          then the DNA string in tw is 'TGAC'
Kmer Kmer::twin() const {
  Kmer km(*this);

  for (size_t i = 0; i < k_bytes; i++) {
    km.bytes[i] = ~bytes[i];
  }

  km.bytes[k_bytes-1] ^= ~k_modmask;
  km.shiftForward(8*k_bytes-2*k);
  uint8_t tmp;

  for (size_t i = 0; i < k_bytes/2; ++i) {
    tmp = km.bytes[i];
    km.bytes[i] = base_swap[km.bytes[k_bytes-1-i]];
    km.bytes[k_bytes-1-i] = base_swap[tmp];
  }

  if ((k_bytes %2) == 1) {
    km.bytes[k_bytes/2] = base_swap[km.bytes[k_bytes/2]];
  }

  return km;
}


// use:  link = km.getLink(index);
// pre:  0 <= index < 8
// post: gives the forward kmer with the (index % 4) character in 'A','C','G' or 'T' if index < 4
//       else the backward kmer with the (index % 4) character in 'A','C','G' or 'T'
Kmer Kmer::getLink(const size_t index) const {
  assert(index >= 0 && index < 8);
  char c;

  switch (index % 4) {
  case 0: c = 'A'; break;
  case 1: c = 'C'; break;
  case 2: c = 'G'; break;
  case 3: c = 'T'; break;
  }

  return (index < 4) ? forwardBase(c) : backwardBase(c);
}


// use:  fw = km.forwardBase(c)
// pre:
// post: fw is the forward kmer from km with last character c,
//       i.e. if the DNA string in km is 'ACGT' and c equals 'T' then
//       the DNA string in fw is 'CGTT'
Kmer Kmer::forwardBase(const char b) const {
  int s = 2*((k+3) % 4);
  Kmer km(*this);
  km.shiftBackward(2);
  km.bytes[k_bytes-1] &= Kmer::k_modmask;

  switch(b) {
  case 'A': km.bytes[k_bytes-1] |= 0x00 << s; break;
  case 'C': km.bytes[k_bytes-1] |= 0x01 << s; break;
  case 'G': km.bytes[k_bytes-1] |= 0x02 << s; break;
  case 'T': km.bytes[k_bytes-1] |= 0x03 << s; break;
  }

  return km;
}


// use:  bw = km.backwardBase(c)
// pre:
// post: bw is the backward kmer from km with first character c,
//       i.e. if the DNA string in km is 'ACGT' and c equals 'T' then
//       the DNA string in bw is 'TACG'
Kmer Kmer::backwardBase(const char b) const {
  Kmer km(*this);
  km.shiftForward(2);
  km.bytes[k_bytes-1] &= Kmer::k_modmask;

  if (k%4 == 0 and k_bytes < MAX_K/4) {
    km.bytes[k_bytes] = 0x00;
  }

  switch(b) {
  case 'A': km.bytes[0] |= 0x00; break;
  case 'C': km.bytes[0] |= 0x01; break;
  case 'G': km.bytes[0] |= 0x02; break;
  case 'T': km.bytes[0] |= 0x03; break;
  }

  return km;
}


// use:  km.printBinary();
// pre:
// post: The bits in the binary representation of the
//       DNA string for km has been printed to stdout
void Kmer::printBinary() const {
  char buff[9]; buff[8] = '\0';
  printf("binary:");

  for (size_t i = 0; i < Kmer::k_bytes; i++) {
    int2bin(bytes[i],buff,8);
    printf("%s",buff);
  }

  printf("\n");
}


// use:  km.toString(s);
// pre:  s has space for k+1 elements
// post: s[0,...,k-1] is the DNA string for the Kmer km and s[k] = '\0'
void Kmer::toString(char *s) const {
  size_t i,j,l;

  for (i = 0; i < k; i++) {
    j = i % 4;
    l = i / 4;

    switch(((bytes[l]) >> (2*j) )& 0x03 ) {
    case 0x00: *s = 'A'; ++s; break;
    case 0x01: *s = 'C'; ++s; break;
    case 0x02: *s = 'G'; ++s; break;
    case 0x03: *s = 'T'; ++s; break;
    }
  }

  *s = '\0';
}

std::string Kmer::toString() const {
  char buf[MAX_K];
  toString(buf);
  return std::string(buf);
}



// use:  km.shiftForward(i);
// pre:  i = 2,4,6
// post: The DNA string in km has been shifted i/2 positions forward i.e.
//       if i=2 then ACGT becomes XACG and X is A,C,G or T
void Kmer::shiftForward(int shift) {
  if (shift>0) {
    if (shift < 8 ) {
      for (size_t i = Kmer::k_bytes-1; i > 0; i--) {
        bytes[i] <<= shift;
        bytes[i] |= (uint8_t) (bytes[i-1] >> (8-shift));
      }

      bytes[0] <<= shift;
    } else {
      assert(0); // we should never need this!
    }
  }
}


// use:  km.shiftBackward(i);
// pre:  i = 2,4,6
// post: The DNA string in km has been shifted i/2 positions backward i.e.
//       if i=2 then ACGT becomes CGTX and X is A,C,G or T
void Kmer::shiftBackward(int shift) {
  if (shift > 0) {
    if (shift < 8) {
      for (size_t i = 0; i < Kmer::k_bytes-1; i++) {
        bytes[i] >>= shift;
        bytes[i] |= (uint8_t) ( bytes[i+1] << (8-shift));
      }

      bytes[Kmer::k_bytes-1] >>= shift;
    } else {
      assert(0); // bad
    }
  }
}


// use:  set_k(k);
// pre:  this method has not been called before and 0 < k < MAX_K
// post: The Kmer size has been set to k
void Kmer::set_k(unsigned int _k) {
  assert(_k < MAX_K);
  assert(_k > 0);
  assert(k_bytes == 0); // we can only call this once
  k = _k;
  k_bytes = (_k+3)/4;
  //  k_longs = (_k+15)/16;
  k_modmask = (1 << (2*((k%4)?k%4:4)) )-1;
}


unsigned int Kmer::k = 0;
unsigned int Kmer::k_bytes = 0;
//unsigned int Kmer::k_longs = 0;
unsigned int Kmer::k_modmask = 0;