Codes in this repo has been written by Tay Hui Yi for CS4330 - Computational Methods in Bioinformatics, under Prof. Sung Wing Kin.

kmers are substrings of length k contained within biological sequences and are generally composed of nucleotides (A,T,C,G). Generally, the term kmer refers to all of the substrings of length k in a particular DNA sequence (Figure 1). Hence, a sequence of length L will have (L-k+1) kmers. In the context of DNA sequences, the total number of possible kmers is 4^k. In bioinformatics, kmers are mainly used within the context of computational genomics and sequence analysis.

Figure 1. The sequence ATGG has two 3-mers: ATG and TGG.

The aim of the kmer program is to generate the counts of all kmers in a fasta file that occur at least q (user defined) times. A memory-based approach (Counting BloomFilter) is implemented to carry out kmer counting. A new output txtfile containing the kmers that occur at least q times and its respective count is generated.

• Only the canonical form of kmers are counted, where the canonical form is the alphabetically smaller kmer among the kmer itself and its reverse complement.
• kmers containing n or N indicate a missing nucleotide and are disregarded during kmer counting.

1. Ensure that the input fasta files containing the DNA sequence(s) are formatted in the following manner.

    >s1 
    ATCGCTG...
    >s2
    ATCGCTG...

2. Download and unzip the kmer_counting.zip in the location with the input fasta files.
3. Manually change all paths in the main method of kmer.java to the location containing the input fasta files.
4. To run the program, run the following commands in your terminal.

   /*
   Description of parameters 
     f: Name of the input file containing the DNA sequence(s)
     k: length of kmers to be counted 
     q: minimum number of occurences for kmer to be counted 
     
   Example 
     java kmer 15mer.fa 15 3
   */
   
   javac *.java 
   java kmer f k q

5. Open the output txtfile, "f_out.txt", for the generated kmer counts.

• A disk-based approach (DSK algorithm) and accuracy improvements are currently undergoing implementation. The program will be updated in the near future.
• A summary of the methods used in the program can be found under the "Summary.pdf".
• Terminal will output the total RAM and wall-clock running time of each input fasta file.