Bio Sequence / Elal Gilboa

🔧 Extensions:

• EXTQUALITY – Filters out low-quality reads, low-quality Kmers, and highly redundant Kmers during the mapping process.
• EXTREVCOMP – Determines the best orientation for each read (forward/reverse) based on the number of unique Kmers.
• EXTCOVERAGE – Calculates genome coverage based on unique and ambiguous read alignments.
• EXTSIM – Filters highly similar genomes from the reference Kmer DB and provides filtering statistics.

🚀 Usage

Run from the command line:

python3 main.py -t [chosen_task] [chosen_flags]

✅ Valid Flag Combinations:

Reference Creation

-t reference -g genome.fa -k 21 -r output.kdb

Required:

• -g: genome file path (FASTA format)
• -k: kmer size (21–31)
• -r: output reference DB file path (.kdb)

Optional (for EXTENSIM):

• --filter-similar
• --similarity-threshold THRESHOLD

Dump Reference

-t dumpref -r ref.kdb

Or:

-t dumpref -g genome.fa -k 21

Align Reads

-t align -r ref.kdb -a output.aln --reads input.fq [options]

Or:

-t align -g genome.fa -k 21 -a output.aln --reads input.fq [options]

Options:

• -m: unique threshold
• -p: ambiguous threshold

Dump Alignments

-t dumpalign -a output.aln

Can only be used alone.

Or:

-t dumpalign -r ref.kdb --reads input.fq [options]

Or:

-t dumpalign -g genome.fa -k 21 --reads input.fq [options]

🔌 Extension Parameters

• EXTQUALITY: --min-read-quality MRQ, --min-kmer-quality MKQ, --max-genomes MG
• EXTREVCOMP: --reverse-complement
• EXTCOVERAGE: --coverage, --genomes g1,g2,..., --min-coverage MC, --full-coverage

🧠 Design

Data Structures:

• Reference Kmers DB is implemented as a dict[str, List[int]] for O(1) lookup time in pseudo-alignment.

Classes:

• Kmer: holds sequence data, locations, uniqueness, and reference sources.
• Read: holds read data, quality, header, mapping status, and orientation.
• Reference: holds genome data, header, Kmers, and ID for similarity calculations.
• Alignment: manages mapping results, coverage, and quality statistics.
• KmerDB: contains all reference Kmers and optional similarity statistics.

File Formats:

• *.aln: pickled Alignment object, compressed using gzip
• *.kdb: pickled KmerDB and genome lengths dictionary (from FASTA headers)

FASTA/Q Loading:

• FASTQ: reads 4-line blocks and creates Read objects.
• FASTA: reads genome header + sequence to create Reference objects using generators.

---

Developed by Elal Gilboa
Hebrew University – Bioinformatics Project