def score_set(
set_id,
bg_dist_mean,
primers,
chr_ends,
score_fun,
score_expression,
max_fg_bind_dist,
bg_genome_len=None,
interactive=False):
binding_locations = swga.locate.linearize_binding_sites(primers, chr_ends)
max_dist = max(swga.score.seq_diff(binding_locations))
# Abort now if it's not passing filter (and it's not a user-supplied set)
if not interactive and max_dist > max_fg_bind_dist:
return False, max_dist
if not bg_dist_mean and not bg_genome_len:
swga.error("Neither background length nor ratio were provided, "
"cannot calculate bg_dist_mean")
elif not bg_dist_mean:
bg_dist_mean = float(bg_genome_len)/sum(p.bg_freq for p in primers)
set_score, variables = score_fun(
primer_set=primers,
primer_locs=binding_locations,
max_dist=max_dist,
bg_dist_mean=bg_dist_mean)
add_set = True
# If it's user-supplied, they have the option of not adding it to db
if interactive:
set_dict = dict(
{'score': set_score,
'scoring_fn': score_expression}.items() + variables.items())
swga.message("Set statistics:\n - " + "\n - ".join(
fmtkv(k, v) for k, v in set_dict.items()))
if click.confirm("Add set to database?"):
# User-provided sets have negative numbers, so we find the smallest
# and decrement by 1
set_id = Set.select(fn.Min(Set._id)).scalar() - 1
else:
add_set = False
if add_set:
s = swga.database.add_set(
_id=set_id,
primers=primers,
score=set_score,
scoring_fn=score_expression,
**variables)
set_added = s is not None
if interactive and set_added:
swga.message("Set {} added successfully.".format(set_id))
elif interactive:
swga.message("That primer set already exists.")
return set_added, max_dist
def main(argv, cfg_file):
cmd = Command('find_sets', cfg_file=cfg_file)
score_cmd = Command('score', cfg_file=cfg_file)
cmd.parse_args(argv)
score_cmd.parse_args(argv)
init_db(cmd.primer_db)
# We need to clear all the previously-used sets each time due to uniqueness
# constraints
allsets = Set.select()
if allsets.count() > 0:
if not cmd.force:
click.confirm("Remove all previously-found sets?", abort=True)
for s in progress.bar(allsets, expected_size=allsets.count()):
s.primers.clear()
s.delete_instance()
make_graph(cmd.max_dimer_bp, graph_fname)
swga.message("Now finding sets. If nothing appears, try relaxing your parameters.")
if cmd.workers <= 1:
setlines = setfinder.find_sets(
cmd.min_bg_bind_dist,
cmd.min_size,
cmd.max_size,
cmd.bg_genome_len,
graph_fp=graph_fname)
else:
setlines = setfinder.mp_find_sets(
nprocesses=cmd.workers,
graph_fp=graph_fname,
min_bg_bind_dist=cmd.min_bg_bind_dist,
min_size=cmd.min_size,
max_size=cmd.max_size,
bg_genome_len=cmd.bg_genome_len)
score_sets(
setlines,
cmd.fg_genome_fp,
score_cmd.score_expression,
cmd.max_fg_bind_dist,
cmd.max_sets)
def summary(primer_db, fg_length, bg_length):
db = swga.database.init_db(primer_db)
db.connect()
swga.database.create_tables(drop=False)
avg_fg_bind, avg_bg_bind, nprimers = (
Primer
.select(fn.Avg(Primer.fg_freq),
fn.Avg(Primer.bg_freq),
fn.Count(Primer.seq))
.scalar(as_tuple=True))
if (avg_fg_bind is None) or (avg_bg_bind is None):
raise swga.error(
"Could not calculate summary statistics; database may be corrupt")
fg_bind_ratio = avg_fg_bind / float(fg_length)
bg_bind_ratio = avg_bg_bind / float(bg_length)
nactive = Primer.select().where(Primer.active==True).count()
min_tm, max_tm, avg_tm = (
Primer
.select(fn.Min(Primer.tm),
fn.Max(Primer.tm),
fn.Avg(Primer.tm))
.where(Primer.active==True)
.scalar(as_tuple=True))
nsets = Set.select(fn.Count(Set._id)).scalar()
if nsets > 0:
bs = Set.select().order_by(Set.score).limit(1).get()
bs_primers = ", ".join(swga.database.get_primers_for_set(bs._id)).strip()
best_set = bs._id
bs_size = bs.set_size
bs_score = bs.score
bs_stats = "- "+"\n - ".join(
fmtkv(k, v)
for k, v in bs.__dict__['_data'].items()
if k not in ["_id", "pids", "score"]
)
version_header = (
"---------------------\n"
"==== SWGA v{version} ====\n"
"---------------------\n"
.format(version=swga.__version__)
)
summary_msg = """
{version_header}
PRIMER SUMMARY
---------------
There are {nprimers} primers in the database.
{nactive} are marked as active (i.e., they passed filter steps and will be used to find sets of compatible primers.) {ifzero_primers_msg}
The average number of foreground genome binding sites is {avg_fg_bind:.0f}.
(avg binding / genome_length = {fg_bind_ratio:05f})
The average number of background genome binding sites is {avg_bg_bind:.0f}.
(avg binding / genome_length = {bg_bind_ratio:05f})
{melting_tmp_msg}
SETS SUMMARY
---------------
There are {nsets} sets in the database.
{set_msg}---------------
Report generated from {primer_db}
"""
ifzero_primers_msg = colored.green(
"Run `swga filter` to identify primers to use."
if nactive == 0 else "")
melting_tmp_msg = (
"""The melting temp of the primers ranges between {min_tm:.2f}C and {max_tm:.2f}C with an average of {avg_tm:.2f}C."""
if nactive > 0 and min_tm and max_tm else
"No melting temps have been calculated yet.")
ifzero_sets_msg = colored.green(
"Run `swga find_sets` after identifying valid primers to begin collecting sets.\n")
set_msg = ("""
The best scoring set is #{best_set}, with {bs_size} primers and a score of {bs_score:03f}.\nVarious statistics:
{bs_stats}
The primers in Set {best_set} are:
{bs_primers}
""" if nsets > 0 else ifzero_sets_msg)
primer_db = os.path.abspath(primer_db)
nprimers = colored.blue(nprimers, bold=True)
nactive = colored.blue(nactive, bold=True)
nsets = colored.blue(nsets, bold=True)
set_msg = set_msg.format(**locals())
melting_tmp_msg = melting_tmp_msg.format(**locals())
version_header = colored.green(version_header)
summary_msg = summary_msg.format(**locals())
with indent(2):
puts(max_width(summary_msg, 80))