From: Tom Clegg Date: Wed, 4 Mar 2020 20:31:43 +0000 (-0500) Subject: Add example script. X-Git-Url: https://git.arvados.org/lightning.git/commitdiff_plain/dda02024528eb85b2da7dc783f4922e75abae09b Add example script. Arvados-DCO-1.1-Signed-off-by: Tom Clegg --- diff --git a/example-1kg-pca.py b/example-1kg-pca.py deleted file mode 100755 index 4e715d016e..0000000000 --- a/example-1kg-pca.py +++ /dev/null @@ -1,19 +0,0 @@ -#!/usr/bin/env python - -""" -lightning gvcf2numpy -tag-library ~/keep/by_id/su92l-4zz18-92bx4zjg5hgs3yc/tagset.fa.gz -ref ./hg38.fa.gz ~/keep/by_id/su92l-4zz18-s3e6as6uzsoocsb > example.npy -example-1k-pca.py example.npy -example-1k-plot.py example.npy.pca.npy sample_info.csv ~/keep/by_id/su92l-4zz18-s3e6as6uzsoocsb -ls -l example.npy.pca.npy.png -""" - -import sys -infile = sys.argv[1] - -import scipy -X = scipy.load(infile) - -from sklearn.decomposition import PCA -pca = PCA(n_components=4) -X = pca.fit_transform(X) -scipy.save(infile+".pca.npy", X) diff --git a/example-1kg-plot.py b/example-1kg-plot.py deleted file mode 100755 index dcfce784c3..0000000000 --- a/example-1kg-plot.py +++ /dev/null @@ -1,65 +0,0 @@ -#!/usr/bin/env python - -import csv -import os -import scipy -import sys - -infile = sys.argv[1] -X = scipy.load(infile) - -colors = None -if len(sys.argv) > 2: - labels = {} - for fnm in os.listdir(sys.argv[3]): - if '.2.fasta' not in fnm: - labels[fnm] = '---' - if len(labels) != len(X): - raise "len(inputdir) != len(inputarray)" - with open(sys.argv[2], 'rb') as csvfile: - for row in csv.reader(csvfile): - ident=row[0] - label=row[1] - for fnm in labels: - if row[0] in fnm: - labels[fnm] = row[1] - colors = [] - labelcolors = { - 'PUR': 'firebrick', - 'CLM': 'firebrick', - 'MXL': 'firebrick', - 'PEL': 'firebrick', - 'TSI': 'green', - 'IBS': 'green', - 'CEU': 'green', - 'GBR': 'green', - 'FIN': 'green', - 'LWK': 'coral', - 'MSL': 'coral', - 'GWD': 'coral', - 'YRI': 'coral', - 'ESN': 'coral', - 'ACB': 'coral', - 'ASW': 'coral', - 'KHV': 'royalblue', - 'CDX': 'royalblue', - 'CHS': 'royalblue', - 'CHB': 'royalblue', - 'JPT': 'royalblue', - 'STU': 'blueviolet', - 'ITU': 'blueviolet', - 'BEB': 'blueviolet', - 'GIH': 'blueviolet', - 'PJL': 'blueviolet', - } - for fnm in sorted(labels.keys()): - colors.append(labelcolors[labels[fnm]]) - -from matplotlib.figure import Figure -from matplotlib.patches import Polygon -from matplotlib.backends.backend_agg import FigureCanvasAgg -fig = Figure() -ax = fig.add_subplot(111) -ax.scatter(X[:,0], X[:,1], c=colors, s=60, marker='o', alpha=0.5) -canvas = FigureCanvasAgg(fig) -canvas.print_figure(infile+".png", dpi=80) diff --git a/example-su92l-1kg.sh b/example-su92l-1kg.sh new file mode 100644 index 0000000000..147585cac1 --- /dev/null +++ b/example-su92l-1kg.sh @@ -0,0 +1,24 @@ +#!/bin/bash + +go run . build-docker-image +arv keep docker lightning-runtime + +project=su92l-j7d0g-jzei0m9yvgauhjf +gvcf=su92l-4zz18-ykpcoea5nisz74f +fasta=su92l-4zz18-s3e6as6uzsoocsb +tags=su92l-4zz18-92bx4zjg5hgs3yc + +go run . import -project ${project} \ + -tag-library ~/keep/by_id/${tags}/tagset.fa.gz \ + ~/keep/by_id/${fasta} +go run . filter -project ${project} \ + -i ~/keep/by_id/su92l-4zz18-fcyucnod8y4515p/library.gob \ + -min-coverage 0.9 -max-variants 30 +go run . export-numpy -project ${project} \ + -i ~/keep/by_id/su92l-4zz18-l40xcd2l6dmphaj/library.gob +go run . pca -project ${project} \ + -i ~/keep/by_id/su92l-4zz18-i6fzfoxpdh38yk4/library.npy +go run . plot -project ${project} \ + -i ~/keep/by_id/su92l-4zz18-zqfo7qc3tadh6zb/pca.npy \ + -labels-csv ~/keep/by_id/${gvcf}/sample_info.csv \ + -sample-fasta-dir ~/keep/by_id/${fasta}