anthony212/auto_blastp repository overview

This container is built using a combination of Ubuntu and Python to automatically BLAST large collections of protein sequences. It collects and returns the list of species that match the BLAST results.

It can take any size of input sequences, but only BLASTs 5 sequences at a time as to not overwhelm NCBI's shared servers, and waits for the completion of those 5 searches before proceeding with the next set. Because of the long running nature of this process it is advisable to use --restart=on-failure so the container will be restarted in the event of an error.

⁠Usage

docker create \
  --name auto_blast \
  -v </path/to/your/data/folder>:/data \
  --restart=on-failure \
  anthony212/auto_blastp

docker start auto_blast

⁠Parameters

-v /data - folder where data is read from and written to by the container.

This parameter is split into two halves, separated by a colon (host:container). First part is your local folder containing the data, second part is the containers folder /data. An example would be -v /home/user/host_data:/data. Make sure not to include any / at the end of the paths.

⁠Input Data Requirements

The input data must be correctly formatted to be read correctly by the container. It must be in comma-separated format. This can be achieved by creating the table in Excel and saving it as a .csv file.

The filename must be data.csv and located in the folder pointed to in -v

Columns should be as such and in this order: id,clone,gene,sequence

id - a unique integer identifying the row clone - string identify this row, helpful if unique, but not required gene - gene name identifying this row sequence - a protein sequence using 1 letter abbreviations that should be BLASTed

Presently this program uses the default BLAST parameters for searching protein sequences. However, if you are BLASTing short protein sequences (less than 10), results may be missed and this would be indicated by missing data on the output spreadsheets.

⁠Output Data

When started this container loads your input data into a database. This database is continuously updated as BLAST searches execute and complete. This enables the search process to be resumed if the program is interrupted or restarted. However if this interruption is more than 24 hours, data may be missed as NCBI BLAST only retains search data for a finite period of time.

This database file is saved into your input data folder as antibody.db. This search can be resumed as long as this file remains in the folder. If this file is removed. The BLAST search will start from the beginning and create a new database from scratch.

⁠Source

Source code is available on GitHub⁠