Experiments and data analysis by Mike Doud, Juhye Lee, and Jesse Bloom. The analysis uses the dms_tools2 software package.
This is analysis for the study published in Nature Communications, 9:1386.
We performed mutational antigenic profiling of three broadly neutralizing monoclonal antibodies, FI6v3, S139/1, and C179, that target influenza hemagglutinin. We used mutant virus libraries of the A/WSN/1933 (H1N1) HA using the procedure described in Doud, Hensley, and Bloom (2017). We are also including analysis of the monoclonal antibodies H17-L19, H17-L10, and H17-L7, previously profiled in Doud, Hensley, and Bloom (2017).
We then analyzed the data both to determine the antigenic effects of all mutations with respect to each antibody, and to compare the overall ease of escape across antibodies.
We selected three biological replicate mutant virus libraries carrying mutant HA's with different neutralizing concentrations of FI6v3, S139/1, C179, or one of the three H17-L* antibodies (performed previously). We then infected cells with the neutralized viruses, extracted RNA from the infected cells, reverse-transcribed and amplified the extracted RNA, and deep sequenced the libraries using a barcoded-subamplicon approach.
In this analysis, we examine deep sequencing data from these antibody selections and evaluate amino-acid mutations that enable escape from each of these six antibodies.
In [1]:
import os
import glob
import math
import time
import pandas
import numpy
from IPython.display import display, HTML
import dms_tools2
import dms_tools2.plot
import dms_tools2.sra
import dms_tools2.diffsel
import dms_tools2.fracsurvive
from dms_tools2.ipython_utils import showPDF
print('Using dms_tools2 version {0}'.format(dms_tools2.__version__))
# define created directories
resultsdir = './results/'
if not os.path.isdir(resultsdir):
os.mkdir(resultsdir)
fastqdir = os.path.join(resultsdir, 'FASTQ_files/')
if not os.path.isdir(fastqdir):
os.mkdir(fastqdir)
countsdir = os.path.join(resultsdir, 'codoncounts/')
if not os.path.isdir(countsdir):
os.mkdir(countsdir)
renumberedcountsdir = os.path.join(resultsdir, 'renumberedcounts')
if not os.path.isdir(renumberedcountsdir):
os.mkdir(renumberedcountsdir)
fracsurvivedir = os.path.join(resultsdir, 'fracsurvive/')
if not os.path.isdir(fracsurvivedir):
os.mkdir(fracsurvivedir)
fracsurviveaboveavgdir = os.path.join(resultsdir, 'fracsurviveaboveavg/')
if not os.path.isdir(fracsurviveaboveavgdir):
os.mkdir(fracsurviveaboveavgdir)
prefsdir = os.path.join(resultsdir, 'prefs')
if not os.path.isdir(prefsdir):
os.mkdir(prefsdir)
# CPUs to use, should not exceed the number you request with slurm
ncpus = 4
# do we use existing results or generate everything new?
use_existing = 'yes'
As input data, we have a CSV file (data/samples.csv) that defines the samples. We read this file into a pandas dataframe that lists the sample names with the following associated information:
none for mock selections and WT plasmid control)
In [2]:
samples = pandas.read_csv('./data/samples.csv')
display(HTML(samples.to_html(index=False)))
FASTQ files from the SRAAll of the FASTQ files are on the Sequence Read Archive (SRA) under the run numbers listed in the samples dataframe defined above.
To download these files, we just pass that dataframe to the dms_tools2.sra.fastqFromSRA function.
Note that this requires the fastq-dump and aspera programs to be installed on the computer you are using at the specified paths.
In [3]:
print('Downloading FASTQ files from the SRA...')
dms_tools2.sra.fastqFromSRA(
samples=samples,
fastq_dump='fastq-dump', # valid path to this program on the Hutch server
fastqdir=fastqdir,
aspera=(
'/app/aspera-connect/3.5.1/bin/ascp', # valid path to ascp on Hutch server
'/app/aspera-connect/3.5.1/etc/asperaweb_id_dsa.openssh' # Aspera key on Hutch server
),
overwrite={'no':True, 'yes':False}[use_existing],
)
print('Completed download of FASTQ files from the SRA')
We used a barcoded-subamplicon sequencing approach to prep and deep sequence the libraries to high accuracy, as described in Doud, Hensley, and Bloom (2017). We will now align the deep sequencing reads using dms2_batch_bcsubamp.
It is also worth noting that we have specified trimming parameters for Read 1 and Read 2. Because different sets of samples were sequenced on different Illumina runs, these trimming parameters also differ across the samples. The trimming was designed to clip off the worst-quality regions of each read for each sample.
We align to the reference sequence in data/WSN_HA_reference.fa.
In [4]:
# file containing wildtype WSN HA sequence
refseq = './data/WSN_HA_reference.fa'
# define subamplicon alignment specifications
alignspecs = ' '.join(['1,285,36,37',
'286,570,31,32',
'571,855,37,32',
'856,1140,31,36',
'1141,1425,29,33',
'1426,1698,40,43'])
# Define trimming parameters
# All S139-selected samples have subamplicon-specific trimming parameters
# The other samples have trimming parameters of 200 for R1trim and 170 for R2trim
R1trim = 200
R2trim = 170
S139_R1trim = ' '.join(['190', '175', '175', '175', '190', '190'])
S139_R2trim = ' '.join(['180', '180', '180', '180', '190', '190'])
samples = (samples
.assign(R1trim=lambda x:
numpy.where(x['name'].str.contains('S139'), S139_R1trim, R1trim),
R2trim=lambda x:
numpy.where(x['name'].str.contains('S139'), S139_R2trim, R2trim)
)
)
# write batch file for dms2_batch_bcsubamp
countsbatchfile = os.path.join(countsdir, 'batch.csv')
samples[['name', 'R1', 'R1trim', 'R2trim']].to_csv(countsbatchfile, index=False)
print('\nNow running dms2_batch_bcsubamp...')
log = !dms2_batch_bcsubamp \
--batchfile {countsbatchfile} \
--refseq {refseq} \
--alignspecs {alignspecs} \
--outdir {countsdir} \
--summaryprefix summary \
--fastqdir {fastqdir} \
--ncpus {ncpus} \
--use_existing {use_existing}
print('Completed dms2_batch_bcsubamp.')
The summary plots from dms2_batch_bcsubamp will have this prefix since it is what was passed as --summaryprefix:
In [5]:
countsplotprefix = os.path.join(countsdir, 'summary')
This plot shows the number of reads per barcode:
In [6]:
showPDF(countsplotprefix + '_readstats.pdf')
This next plot shows the number of reads per barcode. Overall read depth for the C179-associated samples is very good. Some of the S139-selected samples look like they are oversequenced, while others look like they could have been sequenced to greater depth. The two L1mock-A samples also look a little undersequenced.
In [7]:
showPDF(countsplotprefix + '_readsperbc.pdf')
In [8]:
showPDF(countsplotprefix + '_bcstats.pdf')
In [9]:
showPDF(countsplotprefix + '_depth.pdf')
In [10]:
showPDF(countsplotprefix + '_mutfreq.pdf')
In [11]:
showPDF(countsplotprefix + '_codonmuttypes.pdf')
In [12]:
showPDF(countsplotprefix + '_codonntchanges.pdf')
In [13]:
showPDF(countsplotprefix + '_singlentchanges.pdf')
Some of the FI6v3- and H17-L19- selected samples show signs of oxidative damage (enrichment of G to T and C to A mutations). In contrast, the C179- and S139/1- selected samples do not show much signs of oxidative damage. The two sets of mock-selected samples (A and B), which are actually two sequencing reactions of the same initial library prep, do show some signs of oxidative damage.
We will now renumber the codon counts from sequential (1, 2, ... starting from the initial Met) to H3 numbering. The signal peptide is numbered with negative values, and the HA1 and HA2 subunits are numbered separately. This numbering scheme is based on an alignment to a PDB structure for an H3 HA 4HMG. We will use:
The renumbered files will be created in a new renumberedcounts directory and will possess the same names as the original codon counts files created from dms2_batch_bcsubamp above.
In [14]:
renumberfile = './data/H1toH3_renumber.csv'
# counts files to renumber
countsfiles = glob.glob('{0}/*codoncounts.csv'.format(countsdir))
dms_tools2.utils.renumberSites(renumberfile, countsfiles, missing='drop',
outdir=renumberedcountsdir)
We next use dms2_batch_fracsurvive to evaluate the ease of escape from different antibodies. This analysis estimates the fraction surviving antibody treatment for virions with each mutation. Importantly, this metric can be used to compare escape across different antibodies and different concentrations of antibodies.
We first define a fracsurvivebatch dataframe that we write to CSV format to use as input to dms2_batch_fracsurvive.
Note how this dataframe contains a libfracsurvive column that gives the fraction of the library remaining infectious after antibody selection.
This quantity was determined by qRT-PCR of RNA extracted from infected cells after library neutralization.
In [15]:
fracsurvivebatch = pandas.DataFrame.from_records([
# H17L19 at 1 ug/ml
('H17L19-1ug-ml', 'replicate-1a', 'L1-H17L19-1ug-ml-r1', 'L1-mock-r1-A'),
('H17L19-1ug-ml', 'replicate-1b', 'L1-H17L19-1ug-ml-r2', 'L1-mock-r2-A'),
('H17L19-1ug-ml', 'replicate-2', 'L2-H17L19-1ug-ml', 'L2-mock-A'),
('H17L19-1ug-ml', 'replicate-3', 'L3-H17L19-1ug-ml', 'L3-mock-A'),
# H17L19 at 10 ug/ml
('H17L19-10ug-ml', 'replicate-1a', 'L1-H17L19-10ug-ml-r1', 'L1-mock-r1-A'),
('H17L19-10ug-ml', 'replicate-1b', 'L1-H17L19-10ug-ml-r2', 'L1-mock-r2-A'),
('H17L19-10ug-ml', 'replicate-2', 'L2-H17L19-10ug-ml', 'L2-mock-A'),
('H17L19-10ug-ml', 'replicate-3', 'L3-H17L19-10ug-ml', 'L3-mock-A'),
# H17L10 at 3 ug/ml
('H17L10-3ug-ml', 'replicate-1', 'L1-H17L10-3ug-ml', 'L1-mock-r1-A'),
('H17L10-3ug-ml', 'replicate-2', 'L2-H17L10-3ug-ml', 'L2-mock-A'),
('H17L10-3ug-ml', 'replicate-3', 'L3-H17L10-3ug-ml', 'L3-mock-A'),
# H17L7 at 15 ug/ml
('H17L7-15ug-ml', 'replicate-1', 'L1-H17L7-15ug-ml', 'L1-mock-r1-A'),
('H17L7-15ug-ml', 'replicate-2', 'L2-H17L7-15ug-ml', 'L2-mock-A'),
('H17L7-15ug-ml', 'replicate-3', 'L3-H17L7-15ug-ml', 'L3-mock-A'),
# FI6v3 at 0.1 ug/ml (100 ng/ml)
('FI6v3-100ng-ml', 'replicate-1a', 'L1-FI6v3-100ng-ml-r1', 'L1-mock-r1-A'),
('FI6v3-100ng-ml', 'replicate-1b', 'L1-FI6v3-100ng-ml-r2', 'L1-mock-r2-A'),
('FI6v3-100ng-ml', 'replicate-2', 'L2-FI6v3-100ng-ml', 'L2-mock-A'),
('FI6v3-100ng-ml', 'replicate-3', 'L3-FI6v3-100ng-ml', 'L3-mock-A'),
# FI6v3 at 0.2 ug/ml (200 ng/ml)
('FI6v3-200ng-ml', 'replicate-1a', 'L1-FI6v3-200ng-ml-r1', 'L1-mock-r1-A'),
('FI6v3-200ng-ml', 'replicate-1b', 'L1-FI6v3-200ng-ml-r2', 'L1-mock-r2-A'),
('FI6v3-200ng-ml', 'replicate-2', 'L2-FI6v3-200ng-ml', 'L2-mock-A'),
('FI6v3-200ng-ml', 'replicate-3', 'L3-FI6v3-200ng-ml', 'L3-mock-A'),
# C179 at 1 ug/ml
('C179-1ug-ml', 'replicate-1a', 'L1-C179-1ug-ml-r1', 'L1-mock-r1-B'),
('C179-1ug-ml', 'replicate-1b', 'L1-C179-1ug-ml-r2', 'L1-mock-r1-B'),
('C179-1ug-ml', 'replicate-1c', 'L1-C179-1ug-ml-r3', 'L1-mock-r2-B'),
('C179-1ug-ml', 'replicate-2', 'L2-C179-1ug-ml', 'L2-mock-B'),
('C179-1ug-ml', 'replicate-3', 'L3-C179-1ug-ml', 'L3-mock-B'),
# C179 at 2.5 ug/ml (rounded down to 2 ug/ml)
('C179-2ug-ml', 'replicate-1', 'L1-C179-2ug-ml', 'L1-mock-r1-B'),
('C179-2ug-ml', 'replicate-2', 'L2-C179-2ug-ml', 'L2-mock-B'),
('C179-2ug-ml', 'replicate-3', 'L3-C179-2ug-ml', 'L3-mock-B'),
# S139 at 100 ug/ml
('S139-100ug-ml', 'replicate-1', 'L1-S139-100ug-ml', 'L1-mock-r1-B'),
('S139-100ug-ml', 'replicate-2', 'L2-S139-100ug-ml', 'L2-mock-B'),
('S139-100ug-ml', 'replicate-3', 'L3-S139-100ug-ml', 'L3-mock-B'),
# S139 at 200 ug/ml
('S139-200ug-ml', 'replicate-1', 'L1-S139-200ug-ml', 'L1-mock-r1-B'),
('S139-200ug-ml', 'replicate-2', 'L2-S139-200ug-ml', 'L2-mock-B'),
('S139-200ug-ml', 'replicate-3', 'L3-S139-200ug-ml', 'L3-mock-B'),
# S139 at 300 ug/ml
('S139-300ug-ml', 'replicate-1', 'L1-S139-300ug-ml', 'L1-mock-r1-B'),
('S139-300ug-ml', 'replicate-2', 'L2-S139-300ug-ml', 'L2-mock-B'),
('S139-300ug-ml', 'replicate-3', 'L3-S139-300ug-ml', 'L3-mock-B'),
],
columns=['group', 'name', 'sel', 'mock']
)
# all samples have the same error control
fracsurvivebatch['err'] = 'WTplasmid'
# add the libfracsurvive values from the previously defined `samples` dataframe
fracsurvivebatch = pandas.merge(fracsurvivebatch, samples[['name', 'libfracsurvive']],
left_on = 'sel', right_on= 'name', suffixes=('', '_y')).drop('name_y', axis=1)
# make nicely formatted grouplabels (antibody names) for faceted plots
fracsurvivebatch['grouplabel'] = (fracsurvivebatch['group']
.str.replace('-ml', '/ml)')
.str.replace('-', ' (')
.str.replace('ng', ' ng')
.str.replace('ug', ' $\mu$g')
.str.replace('S139', 'broad anti-RBS antibody: S139/1')
.str.replace('H17', 'narrow anti-head antibody: H17')
.str.replace('C179', 'broad anti-stalk antibody: C179')
.str.replace('FI6v3', 'broad anti-stalk antibody: FI6v3')
.str.replace('C179 \(2', 'C179 (2.5')
)
# add column with antibody name, which is word item in group name
fracsurvivebatch['antibody'] = list(map(lambda g: g.split('-')[0], fracsurvivebatch['group']))
# display and write information
fracsurvivebatchfile = os.path.join(fracsurvivedir, 'batch.csv')
print("Here is the batch input that we write to the CSV file {0}:".format(fracsurvivebatchfile))
display(HTML(fracsurvivebatch.to_html(index=False)))
fracsurvivebatch.to_csv(fracsurvivebatchfile, index=False)
We now run dms2_batch_fracsurvive twice with the following difference:
--aboveavg yes option to compute the fraction surviving for each mutation above the overall library average.Note how the results for these two different runs are output to two different subdirectories.
In [16]:
for (arg_aboveavg, outdir) in [('', fracsurvivedir),
('--aboveavg yes', fracsurviveaboveavgdir)]:
print("\nRunning dms2_batch_fracsurvive {0}and writing output to {1}".format(
{'':'', '--aboveavg yes':'with `--aboveavg yes` '}[arg_aboveavg], outdir))
log = !dms2_batch_fracsurvive \
--summaryprefix summary \
--batchfile {fracsurvivebatchfile} \
--outdir {outdir} \
--indir {renumberedcountsdir} \
--use_existing {use_existing} \
{arg_aboveavg}
print("Completed run.")
The results for the analysis with and without the --aboveavg yes option are extremely similar for the antibodies with strong escape mutations.
However, they are a bit clearer for the antibodies without strong escape mutations when we use the --aboveavg yes option.
So for all plots below, we will show the results for the analysis where we compute the average fraction surviving above average.
First, we define the prefix for the summary plots for the fraction surviving above average:
In [17]:
fracsurviveprefix = os.path.join(fracsurviveaboveavgdir, 'summary_')
First, we just look to see the correlations among the replicates for each antibody / concentration.
We do this looking at the average fraction surviving at each site, averaged over all mutations at that site.
These plots have the suffix avgfracsurvive.corr.pdf.
As seen below, the correlations are very good except for the broadly neutralizing anti-stalk antibodies FI6v3 and C179.
As we will see below, the correlations are worse for those antibodies as there are no strong escape mutants and so there is simply less signal to be seen.
In [18]:
for antibody in fracsurvivebatch['antibody'].unique():
groups = fracsurvivebatch.query('antibody == @antibody')['group'].unique()
plots = [fracsurviveprefix + g + '-avgfracsurvivecorr.pdf' for g in groups]
print("\nReplicate correlations for {0}".format(antibody))
showPDF(plots, width=300 * len(plots))
Now we want to look across the HA sequence to see the fraction surviving above average for each antibody / concentration.
There are two ways that we can summarize replicates across an antibody / concentration: by taking the mean, or by taking the median.
Output plots have been created for both of these.
Here we use the median, as that looks slightly cleaner as it reduces the effects of outliers more.
However, the mean looks very similar (you can see this by looking the created plots that have mean in their name in place of median).
First we look at the fraction escape averaged across all mutations at a site:
In [19]:
showPDF(fracsurviveprefix + 'medianavgfracsurvive.pdf', width=800)
Now we look at the fraction escape for the amino-acid mutation with the maximum effect at each site:
In [20]:
showPDF(fracsurviveprefix + 'medianmaxfracsurvive.pdf', width=800)
When considering the plots above, a couple things are clear:
Now for each antibody, we will take the median across antibody concentrations of the across-replicate medians, and use these to make logo plots for each antibody.
In [21]:
medianfiles = []
medavgsitefiles = []
logoplots = []
for antibody in fracsurvivebatch['antibody'].unique():
print('\nGetting and plotting overall across-concentration median for {0}'.format(antibody))
# list of files
medianfracsurvive_files = glob.glob('{0}*{1}-*medianmutfracsurvive.csv'
.format(fracsurviveprefix, antibody))
# Average across mutation fraction surviving
medianmutdf = dms_tools2.fracsurvive.avgMutFracSurvive(
medianfracsurvive_files, 'median')
# Convert median mutation to median avg site fraction surviving
avgsitedf = dms_tools2.fracsurvive.mutToSiteFracSurvive(
medianmutdf)
# Write median mutation fracsurvive dataframe to csv files
medianfile = os.path.join(fracsurviveaboveavgdir,
'antibody_{0}_median.csv'.format(antibody))
medianfiles.append(medianfile)
print("Writing across-concentration medians to {0}".format(
medianfile))
medianmutdf.to_csv(medianfile, index=False)
# Write median average site fracsurvive dataframe to csv files
avgsitefile = os.path.join(fracsurviveaboveavgdir,
'antibody_{0}_median_avgsite.csv'.format(antibody))
medavgsitefiles.append(avgsitefile)
print("Writing across-concentration site medians to {0}".format(
avgsitefile))
avgsitedf.to_csv(avgsitefile, index=False)
# now make logo plot
# scale bar unit is maximum effect
scaleunit = '{0:.1g}'.format(medianmutdf['mutfracsurvive'].max())
scalelabel = '"fraction surviving = {0}"'.format(scaleunit)
logoplot = os.path.join(fracsurviveaboveavgdir,
'{0}_fracsurvive.pdf'.format(antibody))
logoplots.append(logoplot)
print("Creating logo plot {0} for {1} from {2}".format(
logoplot, antibody, medianfile))
log = !dms2_logoplot \
--fracsurvive {medianfile} \
--name {antibody} \
--outdir {fracsurviveaboveavgdir} \
--numberevery 5 \
--nperline 81 \
--underlay yes \
--overlay1 {medianfile} wildtype wildtype \
--scalebar {scaleunit} {scalelabel} \
--use_existing {use_existing}
showPDF(logoplot)
Now we just repeat the above but do it using the fraction surviving values that are not corrected to be above average. We don't show these plots in the notebook, we just create them.
In [22]:
fracsurviveprefix_notexcess = os.path.join(fracsurvivedir, 'summary_')
medianfiles_notexcess = []
medavgsitefiles_notexcess = []
for antibody in fracsurvivebatch['antibody'].unique():
print('\nGetting and plotting overall across-concentration median for {0}'.format(antibody))
# list of files
medianfracsurvive_files = glob.glob('{0}*{1}-*medianmutfracsurvive.csv'
.format(fracsurviveprefix_notexcess, antibody))
# Average across mutation fraction surviving
medianmutdf = dms_tools2.fracsurvive.avgMutFracSurvive(
medianfracsurvive_files, 'median')
# Convert median mutation to median avg site fraction surviving
avgsitedf = dms_tools2.fracsurvive.mutToSiteFracSurvive(
medianmutdf)
# Write median mutation fracsurvive dataframe to csv files
medianfile = os.path.join(fracsurvivedir,
'antibody_{0}_median.csv'.format(antibody))
medianfiles_notexcess.append(medianfile)
print("Writing across-concentration medians to {0}".format(
medianfile))
medianmutdf.to_csv(medianfile, index=False)
# Write median average site fracsurvive dataframe to csv files
avgsitefile = os.path.join(fracsurvivedir,
'antibody_{0}_median_avgsite.csv'.format(antibody))
medavgsitefiles_notexcess.append(avgsitefile)
print("Writing across-concentration site medians to {0}".format(
avgsitefile))
avgsitedf.to_csv(avgsitefile, index=False)
# now make logo plot
# scale bar unit is maximum effect
scaleunit = '{0:.1g}'.format(medianmutdf['mutfracsurvive'].max())
scalelabel = '"fraction surviving = {0}"'.format(scaleunit)
logoplot = os.path.join(fracsurvivedir,
'{0}_fracsurvive.pdf'.format(antibody))
print("Creating logo plot {0} for {1} from {2}".format(
logoplot, antibody, medianfile))
log = !dms2_logoplot \
--fracsurvive {medianfile} \
--name {antibody} \
--outdir {fracsurvivedir} \
--numberevery 5 \
--nperline 81 \
--underlay yes \
--overlay1 {medianfile} wildtype wildtype \
--scalebar {scaleunit} {scalelabel} \
--use_existing {use_existing}
The logoplots above are each dynamically scaled according to the range of the data. For the broad antibodies, we also want to make logoplots that share the same scale. These will be used for some of the figures.
Specifically, we make logo plots for FI6v3 and C179 that share a maximum, and one for S139/1 that has 10 times that maximum.
In [23]:
fracsurvivemax = 0.4 # for C179 / FI6v3
scaleheight = 0.1 # scale bar height for C179 / FI6v3
s139factor = 10 # max for S139/1 this much larger
for f in medianfiles:
if ('C179' in f) or ('FI6v3' in f):
abmax = fracsurvivemax
abscaleheight = scaleheight
elif 'S139' in f:
abmax = s139factor * fracsurvivemax
abscaleheight = s139factor * abscaleheight
else:
continue
antibody = f.split('_')[1]
abname = antibody + '-scaled'
logoplot = os.path.join(fracsurviveaboveavgdir, '{0}_fracsurvive.pdf'.format(abname))
scalelabel = '"fraction surviving = {0:.1f}"'.format(abscaleheight)
print("Creating logo plot {0} for {1} from {2}".format(logoplot, antibody, f))
log = !dms2_logoplot \
--fracsurvive {f} \
--name {abname} \
--outdir {fracsurviveaboveavgdir} \
--numberevery 5 \
--nperline 81 \
--underlay yes \
--overlay1 {f} wildtype wildtype \
--fracsurvivemax {abmax} \
--scalebar {abscaleheight} {scalelabel} \
--use_existing no
We want to see how the sites of escape mutations relate to the inherent mutational tolerance of HA. This inherent mutational tolerance was measured by Doud and Bloom (2016), who determined the preference of each site in HA for each possible amino acid. We will make a logoplot visualizing these preferences. We will use the rescaled preferences averaged across all replicates; these preferences are in H3 numbering. These preferences are taken from the original publication, and are in the file ./data/Overall-WSNHA_merged_prefs_rescaled_H3numbering.csv.
In [24]:
# Define preferences file
prefsfile = './data/Overall-WSNHA_merged_prefs_rescaled_H3numbering.csv'
logoname = 'WSNprefs-H3numbering'
log = !dms2_logoplot \
--prefs {prefsfile} \
--name {logoname} \
--outdir {prefsdir} \
--nperline 81 \
--use_existing no
logoplot = os.path.join(prefsdir, '{0}_prefs.pdf'.format(logoname))
showPDF(logoplot)