Method for diagnosing cancer and predicting type of cancer based on single nucleotide variant in cell-free dna

An apparatus for diagnosing cancer and predicting a type of cancer based on artificial intelligence, comprising:

• a decoding unit configured to decode sequence information by extracting nucleic acids from a biological sample;
• an alignment unit configured to align the decoded sequence to a reference genome database;
• a mutation detection unit configured to extract cancer-specific single nucleotide variants by detecting single nucleotide variants in the aligned sequence and performing filtering;
• a single nucleotide variant distribution calculation unit configured to divide a reference genome into predetermined bins and calculate a regional mutation density of extracted single nucleotide variants in each bin;
• a mutation frequency calculation unit configured to calculate a frequency of mutation signature of the extracted single nucleotide variants;
• a cancer diagnosis unit configured to determine whether cancer is present or not by inputting the calculated regional mutation density of single nucleotide variants and the calculated mutation frequency thereof into an artificial intelligence model trained to perform cancer diagnosis and comparing an output value with a reference value;
• and
• a cancer-type prediction unit configured to predict a type of cancer by inputting the regional mutation density of single nucleotide variants of the sample determined to be cancer and the mutation frequency thereof to a second artificial intelligence model trained to classify types of cancer and comparing output values.

A computer-readable storage medium comprising instructions configured to be executed by a processor for diagnosing cancer and predicting a type of cancer, by:

• (a) extracting nucleic acids from a biological sample to obtain sequence information;
• (b) aligning the obtained sequence information (reads) to a reference genome database;
• (c) extracting cancer-specific single nucleotide variants by detecting single nucleotide variants in the aligned sequence information (reads) and performing filtering;
• (d) dividing the reference genome into predetermined bins and calculating a regional mutation density of extracted single nucleotide variants in each bin;
• (e) calculating a frequency of mutation signature of the extracted single nucleotide variants;
• (f) determining whether cancer is present or not by inputting the regional mutation density of single nucleotide variants calculated in step (d) and the frequency of mutation signature of single nucleotide variants calculated in step (e) into an artificial intelligence model trained to perform cancer diagnosis and comparing an output value with a reference value;
• and
• (g) predicting a type of cancer by inputting the regional mutation density of single nucleotide variants and the frequency of mutation signature of single nucleotide variants of the sample determined to be cancer in step (f) into a second artificial intelligence model trained to classify types of cancer and comparing output values.