Workshop 4 Background Slides

Comparative Genomics: Key Concepts

• Comparing genomes reveals similarities and differences across individuals or populations
• Batch analysis enables efficient, large-scale comparisons

Batch Processing with Snakemake

• Snakemake can automate repetitive tasks across many samples
• Use wildcards and config files to generalize rules
• Modular workflows improve reproducibility and scalability

Generalizing Rules with Wildcards

• Wildcards define patterns in file names
• Defined with curly braces: "{sample}.txt"
• Strings with wildcards are provided to input and output cluases in rule definitions

Wildcard example

rule calculate_gc:
    input:
        "data/{sample}.fasta"
    output:
        "results/{sample}_gc.txt"
    shell:
        "python scripts/gc_content.py {input} > {output}"

{sample} is a wildcard that is defined in the rule definition.

rule all:
    input:
        "results/sampleA_gc.txt"

rule calculate_gc:
    input:
        "data/{sample}.fasta"
    output:
        "results/{sample}_gc.txt"
    shell:
        "python scripts/gc_content.py {input} > {output}"

Snakemake will run:

python scripts/gc_content.py data/sampleA_gc.txt > \
    results/sampleA_gc.txt

rule all:
    input:
        "results/sampleA_gc.txt",
        "results/sampleB_gc.txt",
        "results/sampleC_gc.txt"

rule calculate_gc:
    input:
        "data/{sample}.fasta"
    output:
        "results/{sample}_gc.txt"
    shell:
        "python scripts/gc_content.py {input} > {output}"

Snakemake will run the rule three times, once for each sample in the input of rule all.

The expand() function

• Snakemake defines a function called expand() that can be used to generate a list of files based on a pattern
• It accepts two arguments:
  • A string with one or more wildcard patterns
  • Named arguments for each wildcard that accept lists of values

expand() example

samples = ["sampleA", "sampleB", "sampleC"]
expand("results/{sample}_gc.txt", sample=samples)

produces a list of strings:

["results/sampleA_gc.txt",
 "results/sampleB_gc.txt",
 "results/sampleC_gc.txt"]

Complete example

samples = ["sampleA", "sampleB", "sampleC"]
rule all:
    input:
        expand("results/{sample}_gc.txt", sample=samples)

rule calculate_gc:
    input:
        "data/{sample}.fasta"
    output:
        "results/{sample}_gc.txt"
    shell:
        "python scripts/gc_content.py {input} > {output}"

Running multiple jobs

$ ls
data scripts Snakefile
$ snakemake -j 3 # ← run all three jobs concurrently

Genome Statistics

Basic Genomic Characteristics

• GC content
• Sequence length
• k-mer frequency

Genome Complexity

Generating Comparative Plots

• Visualizations (bar plots, boxplots) make patterns and differences clear
• Plots can summarize GC content, sequence length, and more

Using LLMs for Batch Analysis

• LLMs can help write loops, batch rules, and plotting code
• Effective prompts describe the batch task and desired outputs
• Always review and test generated code for correctness