wormlabcaltech / alaska Goto Github PK
View Code? Open in Web Editor NEWAutomated and friendly RNA-seq analysis (deprecated)
Home Page: http://alaska.caltech.edu
License: MIT License
Automated and friendly RNA-seq analysis (deprecated)
Home Page: http://alaska.caltech.edu
License: MIT License
Provide the users with some kind of visual feedback when they press "Save and apply changes" at each of the metadata cards.
bsdtar
and gnutar
, as indicated at https://apple.stackexchange.com/questions/197839/why-is-extracting-this-tgz-throwing-an-error-on-my-mac-but-not-on-linux.Currently the organism metadata field just says organism. In fact, this should be two fields:
Organism Species: Dropdown menu of scientific names of species
Species Genome: Dropdown menu of the genome versions, ordered in reverse numeric order. I.e., latest genome version first.
Current options for tissue are:
Whole-worm (multi-worm)
Whole-worm (single-worm)
Other
Maybe we should expand other to:
single cell
single cell type/Tissue
For cell type/Tissue, we could provide the largest 10 tissues in C. elegans. Raymond can provide advice on this. If no easy fix, then disregard.
For single purified cells, we should probably offer a dropdown menu of all the neurons, all the early embryonic cells (up to the 8 cell stage), and the distal tip cells/linker cell.
Support non-elegans organisms by downloading and placing the appropriate files from wormbase
Life-stages to choose from in the metadata should include:
**** indicate that Raymond should approve these, since some of these life-stages may have different annotations in wormbase.
Single-cell Embryo ***
2-cell Embryo ***
4-cell Embryo ***
Embryo
L1
L1 arrest
L2
L2d
Dauer
L3
L4
Young Adult
Adult
Post-egg-laying adult (unmated) ****
Post-egg-laying adult (mated) ****
Aged Adult (Animals >7 days old) ****
Modify sleuth.R script to match the analysis of diff_exp_analyzer.r.
This may fix some plots not showing up on the shiny web server.
No description is given for what a misc. characteristic is. An example would be helpful.
Currently, every Sleuth web server opened stays on until the Alaska server is shut down, which is far from ideal. There should be a way to detect when there is no longer activity on the Sleuth server (i.e. detect when the user disconnects) and shut the server down automatically after 'x' minutes/hours of inactivity.
If this is not possible, simply shutting down servers after some time is another option.
We should provide a text like this along with the analysis results. Citations should be included for everything.
RNA-seq data was analyzed using Alaska with using the (single, two)-factor
design option. Briefly, Alaska performs quality control using BowTie2, etc, etc,
etc... and outputs a summary report generated using MultiQC. Read quantification
and differential expression analyses of transcripts was performed using Kallisto
(v.XXXX) and Sleuth (v.XXXX). Kallisto was run using the following flags:
LINE for Single End reads:
-b 200, -l (input), -standarddeviationflag (input), -bias
Line for PE reads:
-b 200, -bias, -
.
Reads were aligned using (Species) genome version (version) as provided by
Wormbase.
Differential expression analyses with Sleuth were performed using a (LR test or
Wald Test) corrected for multiple-testing.
If species == C. elegans: Enrichment analysis was performed using the
WormBase Enrichment Suite.
If two-factor design: Alaska performed epistasis analyses as first presented in
(cite hypoxia paper).
For more accurate file validation.
The 1-factor description currently reads:
A 1-factor design contrasts a control sample with a single experimental sample.
However, I'm hearing from beta users this is confusing, and I agree. We should modify this to read:
A 1-factor design finds the differentially expressed transcripts between an experimentally perturbed sample (for example, a mutant strain) and a reference sample (often the wild-type strain). This is the most common experimental design for RNA-seq.
WormBase has PI identifiers. Maybe we should add this as an optional field?
Lab Identifier: Select your lab identifier from the dropdown. If your lab does not have an identifier, make one (link to page) or leave this blank.
When "open sleuth server" button is clicked on the analysis result page, the sleuth server can not be connected. Currently have no idea why it stopped working... Will have to work on debugging. May be an issue with port forwarding on Docker?
There is an issue where Alaska detects that some projects, though they are still in the object, are stale.
It then removes active projects.
An example is here https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSM1561394.
Also the SOFT file specification for RNA-seq here https://www.ncbi.nlm.nih.gov/geo/info/soft-seq.html and information about processed data files here https://www.ncbi.nlm.nih.gov/geo/info/seq.html#processed.
Specifically, we need to add the following labels:
!Sample_data_processing - Description of the data processing steps & software (including versions)
!Sample_supplementary_file - One for every processed data file
!Sample_processed_data_files_format_and_content - One for every processed data file, in tandem with !Sample_supplementary_file, describing the format and content of the file. Unclear how detailed it needs to be. @dangeles?
Library inclusions were accidentally removed in Sleuth.R
on commit 3b66f0f.
Sleuth.R
should include these libraries: sleuth
, optparse
and files
.
Submitted from the feedback form on the WormBase website.
We are using Alaska software to analyse our RNAseq data. We have two bugs to report:
Implement post_quant_analysis for 1-factor designs.
Scripts for 2-factor designs are not yet implemented but should include:
When the server is started with Start.sh
, automatically scan the saves
directory and load the most recent save.
This would be more convenient for the purposes of testing/managing the server (from the server host), rather than running Request.sh every time on a separate terminal.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.