f = open(filename, 'r')
for line in f:
protein_id = line.strip("\n")
plot_title = ("BLAST(e=" + str(evalue) + ") "
+ protein_id +
" (" + method.fname() + ")")
plot_filename = "roc_plot_" + method.name() + "_" + protein_id
b = benchmark.benchmark(protein_id,
golden_standard=method,
search_method=benchmark.Blast(evalue, max_alignments))
b_random = benchmark.benchmark(protein_id,
golden_standard=method,
search_method=benchmark.RandomUniprot())
roc_plot.roc_plot(b, title=plot_title, filename=plot_filename,
random=b_random)
def usage():
print """This is a script that draws roc plots of using BLAST for homology
search compared to a golden standard (GeneOntology, Pfam or SCOP). One filename
is expected as an argument that contains a list of uniprot protein ids to use.
Command line options:
-n --maxalignments
The number maximum number of alignments to get from SCOP. Default is 100.
-e --e-value
The results from blast are filtered by e-value. BLAST results with a e-value
higher than this are thrown away. Default is 0.1.
from roc_plot import roc_plot
from sklearn.datasets import load_breast_cancer
from sklearn.model_selection import train_test_split
from sklearn.tree import DecisionTreeClassifier
X_train, X_test, y_train, y_test = train_test_split(*load_breast_cancer(return_X_y=True))
roc_plot(
DecisionTreeClassifier,
'max_depth',
[1, 2, 3, 4, 5, 9, 12, 16, 20, 25, 30, 35],
X_train, y_train,
X_test, y_test,
fname='depth_roc.html'
)
