1. Data Setup

---

1.1 For Standard Processing

Name of the dataset denoted as ...

• ./data/raw/msa contains reference MSA with name "..._reference.fasta"

• ./data/raw/query contains query file with name "..._query.fasta"

• ./data/raw/placements contains placement files with name "...epa result.jplace"

• ./data/raw/reference_tree contains reference tree with name "....newick"

1.2 For Leave One Out Processing

Each reference MSA sample will be once a query, the rest is used to reestimate the reference tree

Name of the dataset denoted as ...

• ./data/raw/msa contains reference MSA with name "..._reference.fasta"

• ./data/raw/query contains query file with name "..._query.fasta"

• ./data/raw/reference_tree contains reference tree with name "....newick"

Query and reference are the same file in this case

2. Target Calculation

2.1 For Standard Processing

• ./target/placement_processing.py creates final target files in ./data/processed/target

2.2 For Leave One Out Processing

• ./scripts/leave_one_out.py** performs Leave One Out Processing using epa-ng and raxml-ng and returns placements in ./data/processed/loo_results

• ./target/placement_processing_loo.py creates final Leave One Out target file in ./data/processed/target

3. Feature Generation

3.1 k-mer Distance

./features/kmer_processing.py: