You will be able to run these examples on either AWS or GCP. For most of these examples, you will need to write a configuration file. Below we describe how to write your file and the differences between files for GCP and AWS.
View these examples as suggestions. Once you are confident you understand how ElasticBLAST works, you can start modifying the examples. You can use a local FASTA file, change the database or change the formatting options.
In the first part of this section, we provide information that will help you to complete the examples listed here.
It is possible to run these examples using the Cloud Shell as was done with the GCP and AWS quickstarts. On the other hand, there are advantages to using your own hardware or a cloud instance that you have started. Some advantages to using your own hardware or cloud instance are increased disk space and more processing power, allowing you to better make use of ElasticBLAST as part of a pipeline. If you will be using your own hardware or a cloud instance, you should review the Requirements. You should also look at the section below on Providing Credentials.
Below is the configuration file used in the AWS quickstart (its GCP equivalent
is similar). This file contains three sections:
blast. Each section contains a number of configuration variables (key/value
pairs). These are defined in Configuration variables. Here are some changes you
may need or want to make:
You will be able to use the same configuration variables used in the quickstart, assuming the same account. If you will be changing any part of this section, please refer to GCP Configuration or AWS Configuration.
You do not need to change the configuration variables in this section, though you may want to change the number of machines (
num-nodes). In this section, you can also add a
use-preemptible = yeskey/value pair to indicate that you want to use a less expensive preemptible (GCP) or spot (AWS) instance. See Use preemptible nodes for details. ElasticBLAST will select an appropriate machine type with sufficient memory for your database. You may override this feature and specify a
machine-typein this section, but that is not recommended. See Machine type for information on the default machine types and how to select a different machine type.
You will need to edit the configuration variables in this section in order to accomplish your goal. You can provide BLAST+ specific parameters in this section such as the program, database and other BLAST+ command-line parameters. See BLAST Configuration Options for details.
[cloud-provider] aws-region = us-east-1 [cluster] num-nodes = 2 [blast] program = blastp db = refseq_protein queries = s3://elasticblast-test/queries/BDQA01.1.fsa_aa results = s3://elasticblast-YOURNAME/results/BDQA options = -task blastp-fast -evalue 0.01 -outfmt "7 std sskingdoms ssciname"
If you are not using the cloud shell, you will need to provide credentials if you have not already done so.
Read about providing credentials for GCP here.
For AWS, you can configure access via any of the ways listed in the AWS CLI configuration documentation. If working with an AWS EC2 instance, you can also use AWS IAM roles to grant permissions (see also Required IAM Permissions).