---
tags: tutorials
---

[TOC]

# Sourmash GGG298 Tutorial:
## Putting together everything we learned!

### 1. Install packages with Conda
make conda env, install snakemake minimal, install sourmash

conda create -y --name lab9_ggg298 -c conda-forge -c bioconda

conda activate lab9_ggg298

conda install -c bioconda snakemake-minimal

conda install -c bioconda sourmash

sourmash

sourmash returns help menu

an export.yml file can be created that stores the env info:
`conda env export > export.yml`
but we should edit out all of the dependency lines in that file that aren’t sourmash or snakemake, so the contents are:

name: lab9_ggg298
channels:
  - r
  - conda-forge
  - bioconda
  - defaults
dependencies:
  - sourmash=3.2.2=py37h516909a_0
  - snakemake-minimal=5.10.0=py_0
prefix: /home/hehouts/miniconda3/envs/lab9_ggg298

then we can use
`conda env create -f export.yml -n <new name>` to create a new environment from that file.

### 2. Github
Made a github repo called "lab9_ggg298"
Made a github project called "Sourmash_lab9_ggg298"

git clone https://github.com/hehouts/lab9_ggg298

throws the warning that an empty repository was cloned

### 3. Snakemake
Now we are going to make a snakefile!
`nano -ET4 Snakefile`

The first rule we are writing is to download files:

rule download_data:
    output: "data_files/1.fa.gz",
        "data_files/2.fa.gz",
        "data_files/3.fa.gz",
        "data_files/4.fa.gz",
        "data_files/5.fa.gz"

    shell: """
        wget https://osf.io/t5bu6/download -O data_files/1.fa.gz
        wget https://osf.io/ztqx3/download -O data_files/2.fa.gz
        wget https://osf.io/w4ber/download -O data_files/3.fa.gz
        wget https://osf.io/dnyzp/download -O data_files/4.fa.gz
        wget https://osf.io/ajvqk/download -O data_files/5.fa.gz
    """

### 4. Sourmash
#### compute
run sourmash compute on the zipped genomes:

rule sourmash_compute_sigs:
    input: "data_files/{sample}.fa.gz"
    output: "sourmash_sigs/{sample}.fa.gz.sig"

    shell: """
        sourmash compute -k 31 {wildcard.sample}.fa.gz
    """

but we will need a rule all so.

# list out samples
SAMPLES=['data_files/1.fa.gz',
    'data_files/2.fa.gz',
    'data_files/3.fa.gz',
    'data_files/4.fa.gz',
    'data_files/5.fa.gz']

rule all:
    input:
        # create a new filename for every entry in SAMPLES,
        # replacing {name} with each entry.
        expand("{name}.fa.gz.sig", name=SAMPLES),
        
rule sourmash_compute:
    output: "{sample}.sig"
    input: "{sample}"
    shell:
        "sourmash compute -k 31 {wildcards.sample} -o {output}"

#### compare

rule sourmash_compare:
    input: "sourmash_sigs/1.fa.gz.sig",
        "sourmash_sigs/2.fa.gz.sig",
        "sourmash_sigs/3.fa.gz.sig",
        "sourmash_sigs/4.fa.gz.sig",
        "sourmash_sigs/5.fa.gz.sig"
        
    output: "sm_compare_all.cpm",
        "sm_compare_all.csv"
    shell: """
        sourmash compare {input} -o sm_compare_all.cmp
    """

Actions:

Questions:

what is the number on conda installs?
e.g. `conda install -c bioconda/label/cf201901 sourmash`

Thoughts: