Lambda functions that proxies SELECT ... queries to the logan database
and returns the result as a JSON object.
The function is currently deployed here and made public through a function URL.
Endpoint: https://omdmrhz5lb2nrbmodjtm5fxhqq0uevzh.lambda-url.us-east-1.on.aws/
A very basic authentication scheme is in place. You should always set an
Authorization: Bearer 20240516 header on all requests, otherwise you will
get 403 Forbidden errors.
To use, send a POST request with a JSON payload to the function endpoint. The payload should be an object with a single SELECT property set to the SQL query string one needs to be performed against the dabatase.
If the query is succesful you will get back an array with all the rows in the result set. Each row will be an object with keys as column names and values.
If the query is invalid or there is an error while executing it the response should look like:
{"error":"[ERROR MESSAGE]"}
Example:
curl \
-H 'Authorization: Bearer 20240516' \
-H 'Content-Type':'application/json' \
-X POST \
-d '{"SELECT":"* FROM sra LIMIT 8"}' \
"[ENDPOINT]"
Output:
[
{
"acc":"SRR11121871",
"assay_type":"AMPLICON",
"center_name":"VETMEDUNI VIENNA",
"consent":"public",
"experiment":"SRX7758868",
"sample_name":"vmu_te_IFLPS_K14T17",
"instrument":"Illumina MiSeq",
"librarylayout":"PAIRED",
"libraryselection":"PCR",
"librarysource":"METAGENOMIC",
"platform":"ILLUMINA",
"sample_acc":"SRS6178009",
"biosample":"SAMN14143901",
"organism":"bovine gut metagenome",
"sra_study":"SRP250132",
"releasedate":"2021-05-29T00:00:00.000Z",
"bioproject":"PRJNA607677",
"mbytes":17,"avgspotlen":429,"mbases":29,"library_name":"K14T17SARACow 1417",
"biosamplemodel_sam":"Metagenome or environmental",
"collection_date_sam":null,"geo_loc_name_country_calc":"Austria",
"geo_loc_name_country_continent_calc":"Europe",
"geo_loc_name_sam":"Austria: Pottenstein"
},
... and so on ...
]
You can also do a simple GET request with the table name and a query to match its primary key, to quickly get one record from the database.
Example:
curl \
-H 'Authorization: Bearer 20240516' \
"[ENDPOINT]"/sra/SRR11121871
... would return the first object of the previous example.
To learn about the tables that are available, check out table.
- Change the function endpoint to a more memorable (and static) domain.
- Better authentication scheme, if needed It should be trivial to implement API keys.
- GET only works for tables/columns that are explicitly defined. Make this work automatically so it works with all tables without any setup.
- Return rows as an array without the column names save on bandwith.
Retrieves individual contigs from SRA assembly resources on s3://logan-pub.
The function is currently deployed here and made public through a function URL.
Endpoint: https://ag63ar36qwfndxxfc32b3w57oy0wcugc.lambda-url.us-east-1.on.aws/
A very basic authentication scheme is in place. You should always set an
Authorization: Bearer 20240522 header on all requests, otherwise you will
get 403 Forbidden errors.
To use, send a POST request with a JSON payload to the function endpoint. The payload should be an object with two properties:
- SRA: SRA id of the library you want to use
- contig: A string (or array of strings) with the contig ids that you wish to extract
Example:
curl \
-H 'Authorization: Bearer 20240522' \
-H 'Content-Type':'application/json' \
-X POST \
-d '{"SRA":"DRR000001","contig":["DRR000001_1","DRR000001_2"]}' \
"[ENDPOINT]"
Output:
>DRR000001_1 ka:f:17.316 L:+:722:+ L:-:1565:-
AGTGTCTT[...]
>DRR000001_2 ka:f:17.782 L:+:392:- L:+:2311:- L:-:1220:+
CCGGCACT[...]
You can also do a simple GET request with an SRA id and a single contig id, to quickly fetch it from the S3 repository.
Example:
curl \
-H 'Authorization: Bearer 20240522' \
"[ENDPOINT]"/DRR000001/DRR000001_1
... would return the first contig of the previous example.
- Change the function endpoint to a more memorable (and static) domain.
- Use Anton's f2sz
- Reverse complement and other minimal transform functions as long as they do not require external libraries.