Skip to content

Latest commit

 

History

History
72 lines (61 loc) · 6.94 KB

JSON-README.md

File metadata and controls

72 lines (61 loc) · 6.94 KB
Raw

Specifications for Project Data Input and Project JSON

This README details the specifications for how users provide Alaska with their experiment data/information via the portal. All user-input data, along with all project metadata, are saved in JSON format. An example JSON can be found here: example.json. The contents of this JSON will be used as examples to describe each field and their functions. Every project has its own JSON file in its corresponding project directory. From here on, everything is described in relation to the metadata collection form, which the portal asks the users to complete after uploading their raw reads.

JSON Field Descriptions

This section describes each field in the JSON: its purpose, its use and how it should be shown on the portal.

id

  • Every project is assigned a unique identifier (ID) by the server when the project is created via the Request.sh new_proj command.
  • The project ID is the only way the server differentiates projects.
  • The ID itself is not very relevant to the user, but the user should be able to see it (not edit it).

dir, qc_dir, raw_dir, align_dir, diff_dir, temp_dir

  • Specify the paths to directories that are relevant to the analyses. All of these paths are relative from the root folder of the server. Specifically, dir is the path to the project directory, qc_dir the path to the quality control directory, raw_dir to the raw reads directory, align_dir to the read alignment & quantification directory, diff_dir to the differential expression analysis directory and temp_dir to the temporary directory.
  • All of these paths are purely for internal use. The portal should not display these fields.

jobs

  • Contains a list of job ID's that are linked to this project. A job is created for each step of the analysis. A project that has been completely analyzed will most likely have four jobs.
  • Internal use. Should not be displayed on the portal.

raw_reads

  • Dictionary of raw read files with the folder name as the keys and a list of the files in those folders as values.
  • Not displayed on the portal.

chk_md5

  • Dictionary of MD5 hashes corresponding to the read files in the previous field. The folder names as the keys and list of hashes as values.
  • Not displayed on the portal.

samples

  • Dictionary of objects, with the sample ID as the keys and encodings of AlaskaSample objects as the values.
  • Each sample is independent of the other and should be displayed on the portal as such.
  • An explanation of the fields of every sample object is below.
Sample Object Field Descriptions
  • id: Unique identifier (ID) for the sample. Primarily for internal use, but the user should still be able to see it (but not edit it).
  • name: The name for this sample. More human-friendly than the ID. Taken from the folder name that the user uploaded the sample in. Not editable.
  • type: Specifies whether the reads are single- or paired-end. Single-end reads have an integer 1 in this field and paired-end read have an integer 2. Radio buttons to choose between the two.
  • organism: The model organism of the sample. The user selects the organism from a drop down menu. The selections are fetched from organisms/jsons. This folder contains JSON files for each organism. The drop down selections are simply the list of JSON files in this folder without the .json extension.
  • ref_ver: The model organism reference version. The user selects the version from a drop down menu. The possible selections change depending on what model organism the user chose previously. The list of possible reference versions for a given organism is fetched from the refs field of the organism JSON itself, in organisms/jsons. An example organism JSON is here: caenorhabditis_elegans.json.
  • length: The read length. The user is able to input an integer between 50 and 400.
  • stdef: The standard deviation of the read lengths. The user able to input an integer between 20 and 100.
  • boostrap_n: The number of bootstraps to perform. Always set to 200.
  • reads: The raw reads corresponding to this sample. List of strings that are the paths to each read file.
  • chk_md5: The MD5 checksums corresponding to the read files.
  • projects: List of project IDs that are linked to this sample. Internal use only.
  • meta: Other relevant metadata about this sample.
    • title: Nickname/title for this sample, set by the user.
    • contributors: Contributors to generating this sample.
    • source: Where this sample originated.
    • chars: Sample characteristics. This is a dictionary of string key-value pairs that describe this sample. Users can expand the entries freely, but every sample must have at least one field that is shared among all samples in the project. Examples of fields users may include are: strain, tissue, developmental stage, etc.
    • description: Short description of this sample, input by the user.
    • datetime: Date and time this sample was created. Not visible to the user.

design

  • Either 1 or 2. 1 indicates the project is a single-factor design. A single-factor design is when one experimental factor is being tested. 2 indicates the project is a two-factor design. A two-factor design is when two experimental factors and their relationship are being tested.
  • Alaska currently does NOT support two-factor design.

ctrls

  • A dictionary of sample IDs as keys and control characteristics as values. Both are strings.
  • The keys of the dictionary are the control samples (Note that there can be multiple controls).
  • The values of the dictionary are the control characteristic of each sample. The value must match a key in the chars field of the meta field of the sample object. Also, all samples in the project must have all the control characteristic values that appear in this dictionary. For example, if the dictionary is "ASgyci7n": "genotype", "ASb8063z": "life stage", then every sample in the project MUST have both genotype and life stage fields in their respective chars fields in the meta field of their sample objects.
  • For instance, one element of the dictionary may be "ASgyci7n": "genotype", which indicates that the sample with ID ASgyci7n is a control sample, and that this sample is the control for the genotype characteristic.

progress

  • Indicates the progress of this project.
  • Internal use.

meta

  • title: Title of the project. This is a custom title the user calls this project.
  • summary: Brief summary of the project.
  • contributors: List of contributors of the project. It is currently just a list of names. It may be changed into a dictionary containing names and emails.
  • SRA_center_code: SRA center code. Not necessary and user-editable.
  • email: Email of the individual submitting the project. This is how the user is notified that analysis is finished.
  • datetime: The date and time this project was created. For internal use.