Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Shaopeng #50

Merged
merged 2 commits into from
Oct 25, 2023
Merged

Shaopeng #50

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
12db289
add warning on large db for training step, switch suggestion for buil…
ShaopengLiu1 Oct 25, 2023
e245168
fix typo
ShaopengLiu1 Oct 25, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 16 additions & 6 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -85,6 +85,8 @@ The workflow for YACHT is as follows:
2. Preprocess the reference genomes by removing the "too similar" genomes based on `ANI` using the `ani_thresh` parameter
3. Run YACHT to detect the presence of reference genomes in your sample

</br>

### Creating sketches of your reference database genomes

You will need a reference database in the form of [Sourmash](https://sourmash.readthedocs.io/en/latest/) sketches of a collection of microbial genomes. There are a variety of pre-created databases available at: https://sourmash.readthedocs.io/en/latest/databases.html. Our code uses the "Zipfile collection" format, and we suggest using the [GTDB genomic representatives database](https://farm.cse.ucdavis.edu/~ctbrown/sourmash-db/gtdb-rs214/gtdb-rs214-reps.k31.zip):
Expand All @@ -104,18 +106,20 @@ sourmash sketch dna -f -p k=31,scaled=1000,abund --singleton <your multi-FASTA f
If you have a directory of FASTA files, one per genome:

```bash
## Method 1
# cd into the relevant directory
sourmash sketch dna -f -p k=31,scaled=1000,abund *.fasta -o ../training_database.sig.zip
# cd back to YACHT

## Method 2
## Method 1 (suggested)
# put all full paths of FASTA/FASTQ file into a file, one path per line
find <path of foler containg FASTA/FASTQ files> > dataset.csv
sourmash sketch fromfile dataset.csv -p dna,k=31,scaled=1000,abund -o ../training_database.sig.zip
# cd back to YACHT

## Method 2
# cd into the relevant directory
sourmash sketch dna -f -p k=31,scaled=1000,abund *.fasta -o ../training_database.sig.zip
# cd back to YACHT
```

</br>

### Creating sketches of your sample

You will then create a sketch of your sample metagenome, using the same k-mer size and scale factor
Expand Down Expand Up @@ -155,6 +159,12 @@ In the two preceding steps, you will obtain a k-mer sketch file in zip format (i

### Preprocess the reference genomes (Training Step)

##### Warning: the training process is time-consuming on large database

In our benchmark with `GTDB representive genomes`, it takes `15 minutes` using `16 threads, 50GB of MEM` on a system equipped with a `3.5GHz AMD EPYC 7763 64-Core Processor`. The processing time can be significant when executed on GTDB all genomes OR with limited resources. If only part of genomes are needed, one may use `sourmash sig` command to extract signatures of interests only.

</br>

The script `make_training_data_from_sketches.py` extracts the sketches from the Zipfile-format reference database, and then turns them into a form usable by YACHT. In particular, it removes one of any two organisms that have ANI greater than the user-specified threshold as these two organisms are too close to be "distinguishable".

```bash
Expand Down
Loading