def nonmaxima(self):
nonmaxima_img = np.array(self.image_out_grad.copy())
imagea = np.array(self.image_out_angle.copy())
(h, w) = self.image_out_grad.shape
for i in range(h):
for j in range(w):
if (i == 0 or i == h - 1 or j == 0 or j == w - 1):
nonmaxima_img[i][j] = 0
continue
tq = (imagea[i][j]) % 4
if (tq == 0):
if (self.image_out_grad[i, j] <= self.image_out_grad[i,
j - 1]
or self.image_out_grad[i, j] <=
self.image_out_grad[i, j + 1]):
nonmaxima_img[i][j] = 0
if (tq == 1):
if (self.image_out_grad[i, j] <= self.image_out_grad[i - 1,
j + 1]
or self.image_out_grad[i, j] <=
self.image_out_grad[i + 1, j - 1]):
nonmaxima_img[i][j] = 0
if (tq == 2):
if (self.image_out_grad[i, j] <= self.image_out_grad[i - 1,
j]
or self.image_out_grad[i, j] <=
self.image_out_grad[i + 1, j]):
nonmaxima_img[i][j] = 0
if (tq == 3):
if (self.image_out_grad[i, j] <= self.image_out_grad[i - 1,
j - 1]
or self.image_out_grad[i, j] <=
self.image_out_grad[i + 1, j + 1]):
nonmaxima_img[i][j] = 0
if self.completed < 40:
self.completed += 0.1
self.progress.setValue(self.completed)
self.nonmaxima_img = nonmaxima_img
self.progress.setValue(self.completed)
tab_Var = mplt_plot(self.tab, nonmaxima_img)
self.t_index = self.tabWidget.addTab(tab_Var, 'Non-Maximum Supressed')
self.tabWidget.setCurrentIndex(self.t_index)
QMessageBox.about(self, "Status!", "Non-max image Generated!")
# '''Input Window for Max and Min Threshold'''
input_dialog = QtWidgets.QDialog()
input_ui = Threshold.Ui_Dialog()
input_ui.setupUi(input_dialog)
input_dialog.show()
if input_dialog.exec():
thes_l, Thresh_h = input_ui.onOk()
if thes_l and Thresh_h:
thresh = thes_l, Thresh_h
self.h_thres = np.max(thresh) # Value
self.l_thres = np.min(thresh) # Value
self.thres()
def image_process(img, seq_n, NUM_SEGS, IMG_FRACTION):
#cv2.imwrite(str(seq_n) + "_in.jpg", img)
img = cv2.GaussianBlur(img, (13, 13), 0)
imgThreshed = th.thresholding(img)
imgMiddle, colOffset = get_middle(imgThreshed, IMG_FRACTION)
midCentroids, leftCentroids, rightCentroids, frameAtIntersection = row_segment_centers(
imgMiddle, NUM_SEGS, colOffset)
# Draw all
#for i in range(0,len(midCentroids)):
# cv2.circle(img, midCentroids[i], 5, (255,0,0))
#if leftCentroids != None and rightCentroids != None:
# for i in range(0,len(leftCentroids)):
# cv2.circle(img, leftCentroids[i], 3, (0,255,0))
# for i in range(0,len(rightCentroids)):
# cv2.circle(img, rightCentroids[i], 7, (0,0,255))
# cv2.imshow('imageThreshed',imgThreshed)
# cv2.imshow('image', img)
# cv2.imwrite(str(seq_n) + "_th.jpg", imgThreshed)
# cv2.imwrite(str(seq_n) + "_out.jpg", img)
return midCentroids, leftCentroids, rightCentroids, frameAtIntersection
def legacy():
"""
This is the code that I had originall wrote to do this.
This code was really messy, but worked, so I am leaving it here.
It has been rewritten in the above functions, but could be useful to look
at.
"""
## initialize the scope
scope=DGLab_Leica2.DGLab_Leica2()
## set up the
rows=string.ascii_uppercase[1:7]
columns=range(2,12)
sites=range(0,225)
pos_string=open("PositionList_96well_15x15sites.pos").read()
positions=StagePositions.positionList_parser(pos_string)
scope.mmc.enableContinuousFocus(True)
scope.set_illumination(Illuminations.Cy5)
#start_time=time.asctime().replace(" ","_").replace(":","_")
start_time="Thu_Aug_13_14_38_51_2015"
done_list={}
well_counter={}
## if this is a continuation of an old scan, this will add on to the
## existing log file.
if os.path.exists("Scan_%s.txt" % start_time):
completed_fin=open("Scan_%s.txt" % start_time,"r")
for line in completed_fin:
site=line.split(";")[0]
try:
if "No cells of interest" not in line.split(";")[1]:
well=site.split("-")
if well not in well_counter:
well_counter[well]=0
well_counter+=1
except:
pass
completed_fin.close()
status_out=open("Scan_%s.txt" % start_time,"r")
status_string=status_out.read()
status_out.close()
status_out=open("Scan_%s.txt" % start_time,"w")
status_out.write(status_string)
else:
status_out=open("Scan_%s.txt" % start_time,"w")
data_dir="Scan_%s" % start_time
if os.path.isdir(data_dir) is False:
os.mkdir(data_dir)
order=-1
for row in rows:
## flip the order that the columns are gone through
order=order*-1
for column in columns[::order]:
well="%s%s" % (row,column)
if well not in well_counter:well_counter[well]=0
if well_counter[well]<15:
## reset the well count for this new well.
well_count=well_counter[well]
## iterate through the site in this well.
for site in sites: ## Test case
if "%s%s-Site_%s" % (row,column,site) not in done_list.keys():
done_list["%s%s-Site_%s" % (row,column,site)]=True
print ">Looking at %s%s-Site_%s" % (row,column,site)
x,y=positions["%s%s-Site_%s" % (row,column,site)].get_xy()
## take an image to use for a trigger
scope.set_illumination(Illuminations.qCy5)
scope.mmc.setXYPosition("XYStage",x,y)
scope.mmc.snapImage()
img=scope.mmc.getImage()
## show image
plt.imshow(img)
plt.draw()
## analyze this image
cois=Threshold.coi_finder(img,
threshold=1000,
min_area=100,
max_area=1000)
if cois:
do_stack=True
progress_line=">%s%s-Site_%s;cois=%s\n" % (row,column,site,cois)
else:
do_stack=False
progress_line=">%s%s-Site_%s;No cells of interest\n" % (row,column,site)
status_out.write(progress_line)
## run the stack if there was a COI found
if do_stack:
print ">Stack at %s%s-Site_%s" % (row,column,site)
well_count+=1
well_counter[well]+=1
channels=[Illuminations.DIC,
Illuminations.GFP,
Illuminations.qCy5,
Illuminations.DAPI]
scope.mmc.enableContinuousFocus(False)
mda,md=scope.multi_channel_acquisition(channels)
FileOut.mda_save(out_dir=data_dir,
prefix="%s%s-Site_%s" % (row,column,site),
mda=mda,
md=md,
channels=channels)
print "Stack taken at %s%s-Site_%s" % (row,column,site)
## turn AFC back on
scope.mmc.enableContinuousFocus(True)
## move onto the next well if we have acquired enough
if well_count > 15:
break
## break out of the well if you can't focus
if scope.mmc.isContinuousFocusLocked() is False:
time.sleep(5)
if scope.mmc.isContinuousFocusLocked() is False:
print "Couldnt focus 3"
break
## break out of the column if you still cant focus
if scope.mmc.isContinuousFocusLocked() is False:
time.sleep(5)
if scope.mmc.isContinuousFocusLocked() is False:
print "Couldnt focus 2"
break
## break out of the row if you STILL cant focus.
if scope.mmc.isContinuousFocusLocked() is False:
time.sleep(5)
if scope.mmc.isContinuousFocusLocked() is False:
print "Couldnt focus 1"
break
status_out.close()
def plateScanSite(scope,
illumination,
mda_channels,
x,
y,
visible=True,
name="",
stack_height=5,
slice_height=.5,
minAreaCOI=100,
maxAreaCOI=1000,
thresholdCOI=1000):
"""
This is a function to run at each site of a 96well plate scan.
This requires:
-scope
-scan illumination
-MDA illuminations
-x
-y
Optional:
-visibility of scan image during acquisition
-name this is mostly for progress and debugging purposes
This returns:
-site_imaged: 0 or 1 depending if a MDA was taken
-MDA
-MDA_md
"""
scope.set_illumination(illumination)
scope.mmc.setXYPosition("XYStage",x,y)
scope.mmc.snapImage()
img=scope.mmc.getImage()
## show image
if visible is True:
plt.imshow(img)
plt.show()
plt.draw()
# analyze this image
cois=Threshold.coi_finder(img,
threshold=thresholdCOI,
min_area=minAreaCOI,
max_area=maxAreaCOI)
if cois:
do_stack=True
site_imaged=1
print "COIs found in %s, taking a stack" % name
else:
do_stack=False
site_imaged=0
md=None
mda=None
print "No COIs found in %s" % name
if do_stack:
scope.mmc.enableContinuousFocus(False)
mda,md=scope.multi_channel_acquisition(mda_channels,
z_width=stack_height,
z_increments=slice_height)
scope.mmc.enableContinuousFocus(True)
if visible is True: plt.close()
del img
return site_imaged, mda, md
def mm_XG_iRNA_M6A(sequence, species, modification, M_threshold):
M_threshold = Threshold.choiceThreshold(species, modification, M_threshold)
mm_importance_dict_sorted_list_num_M6A = [
1834, 1500, 1549, 1367, 1310, 1459, 5, 1854, 1877, 58, 1843, 498, 42,
71, 1389, 700, 1775, 213, 906, 942, 81, 16, 1429, 888, 1373, 1817,
1501, 218, 124, 1380, 1436, 60, 12, 1536
]
XG_iRNA_mm_M6A = load('model/mm/mm_xgb_M6A_10fold')
# --------------------------------------------------------------
line = sequence # bug
if re.search(r'[^AUCG]', line.strip()):
nota = "Please input the RNA sequence 'AUCG'."
return nota
length = len(line.strip())
if length < 41:
nota = "Please input the RNA sequence at least 51 bases! "
return nota
mm_n_base_41_M6A = 19
mm_n_base_list_M6A_proba = []
mm_M6A_list_num = []
for base in line.strip()[20:-20]:
mm_n_base_41_M6A += 1
file_inter_M6A = open('inter_use/mm_inter_use_M6A.txt', 'w')
if base == 'A':
new_line = line.strip()[mm_n_base_41_M6A - 20:mm_n_base_41_M6A +
21]
# ---f nc dnc tnc tenc pnc nd dnd ncp dpcp onehot dbe-----------------
print(*iMRMFeatureEtraction.nc(new_line, 41),
*iMRMFeatureEtraction.dnc(new_line, 41),
*iMRMFeatureEtraction.tnc(new_line, 41),
*iMRMFeatureEtraction.tenc(new_line, 41),
*iMRMFeatureEtraction.pnc(new_line, 41),
*iMRMFeatureEtraction.nd(new_line, 41),
*iMRMFeatureEtraction.dnd(new_line, 41),
*iMRMFeatureEtraction.ncp(new_line, 41),
*iMRMFeatureEtraction.dpcp(new_line, 41),
*iMRMFeatureEtraction.onehot(new_line, 41),
*iMRMFeatureEtraction.dbe(new_line, 41),
file=file_inter_M6A)
file_inter_M6A.close()
pddtest = pd.read_csv("inter_use/mm_inter_use_M6A.txt",
sep=' ',
header=None)
X_test = pddtest.values[:, [
x - 1 for x in mm_importance_dict_sorted_list_num_M6A
]]
prediction = XG_iRNA_mm_M6A.predict_proba(X_test)
if prediction[0][1] > M_threshold:
mm_n_base_list_M6A_proba.append(prediction[0][1])
mm_M6A_list_num.append(mm_n_base_41_M6A + 1)
if len(mm_M6A_list_num) > 0:
mm_M6A_list_num.reverse()
mm_n_base_list_M6A_proba.reverse()
n = 0
for i in [i - 1 for i in mm_M6A_list_num]:
line = list(line)
line[i] = '<font color=#00FF00>A</font>(' + str(
mm_n_base_list_M6A_proba[n]) + ')'
n += 1
line = str(line).replace(',', '')
line = line.replace('[', '')
line = line.replace(']', '')
line = line.replace(' ', '')
line = line.replace('\'', '')
line = line.replace('\\r', '')
line_pro = line.replace('fontcolor', 'font color')
line = line_pro
pro_list = re.findall('\(.*?\)', line_pro)
for i in pro_list:
line = line.replace(i, '')
return line, line_pro, mm_M6A_list_num
def mm_XG_iRNA_M5C(sequence, species, modification, M_threshold):
M_threshold = Threshold.choiceThreshold(species, modification, M_threshold)
mm_importance_dict_sorted_list_num_M5C = [
16, 48, 1458, 1197, 299, 859, 1359, 1879, 383, 65, 381, 82, 1519
]
XG_iRNA_mm_M5C = load('model/mm/mm_xgb_M5C_10fold')
# --------------------------------------------------------------
line = sequence # bug
if re.search(r'[^AUCG]', line.strip()):
nota = "Please input the RNA sequence 'AUCG'."
return nota
length = len(line.strip())
if length < 41:
nota = "Please input the RNA sequence at least 51 bases! "
return nota
mm_n_base_41_M5C = 19
mm_n_base_list_M5C_proba = []
mm_M5C_list_num = []
for base in line.strip()[20:-20]:
mm_n_base_41_M5C += 1
file_inter_M5C = open('inter_use/mm_inter_use_M5C.txt', 'w')
if base == 'C':
new_line = line.strip()[mm_n_base_41_M5C - 20:mm_n_base_41_M5C +
21]
# ---f nc dnc tnc tenc pnc nd dnd ncp dpcp onehot dbe-----------------
print(*iMRMFeatureEtraction.nc(new_line, 41),
*iMRMFeatureEtraction.dnc(new_line, 41),
*iMRMFeatureEtraction.tnc(new_line, 41),
*iMRMFeatureEtraction.tenc(new_line, 41),
*iMRMFeatureEtraction.pnc(new_line, 41),
*iMRMFeatureEtraction.nd(new_line, 41),
*iMRMFeatureEtraction.dnd(new_line, 41),
*iMRMFeatureEtraction.ncp(new_line, 41),
*iMRMFeatureEtraction.dpcp(new_line, 41),
*iMRMFeatureEtraction.onehot(new_line, 41),
*iMRMFeatureEtraction.dbe(new_line, 41),
file=file_inter_M5C)
file_inter_M5C.close()
pddtest = pd.read_csv("inter_use/mm_inter_use_M5C.txt",
sep=' ',
header=None)
X_test = pddtest.values[:, [
x - 1 for x in mm_importance_dict_sorted_list_num_M5C
]]
prediction = XG_iRNA_mm_M5C.predict_proba(X_test)
if prediction[0][1] > M_threshold:
mm_n_base_list_M5C_proba.append(prediction[0][1])
mm_M5C_list_num.append(mm_n_base_41_M5C + 1)
if len(mm_M5C_list_num) > 0:
mm_M5C_list_num.reverse()
mm_n_base_list_M5C_proba.reverse()
n = 0
for i in [i - 1 for i in mm_M5C_list_num]:
line = list(line)
line[i] = '<font color=#FFA500>C</font>(' + str(
mm_n_base_list_M5C_proba[n]) + ')'
n += 1
line = str(line).replace(',', '')
line = line.replace('[', '')
line = line.replace(']', '')
line = line.replace(' ', '')
line = line.replace('\'', '')
line = line.replace('\\r', '')
line_pro = line.replace('fontcolor', 'font color')
line = line_pro
pro_list = re.findall('\(.*?\)', line_pro)
for i in pro_list:
line = line.replace(i, '')
return line, line_pro, mm_M5C_list_num
def human_XG_iRNA_pse(sequence, species, modification, M_threshold):
# -------------------------------------------------------------
hg_importance_dict_sorted_list_num_Pse = [
16, 22, 1441, 1389, 1692, 300, 661, 406, 1666, 1660, 1432, 1698, 1695,
338, 738, 10, 1680, 808, 1414, 204, 77, 940, 69, 1408, 74, 980, 31,
234, 8, 292, 178, 1382, 1458, 1, 1689, 13, 1373, 39, 237, 1369, 1444,
245, 66, 54, 1434, 1417, 1396
]
XG_iRNA_hg_Pse = load('model/hg/hg_xgb_Pse_10fold')
# --------------------------------------------------------------
line = sequence
hg_n_base_21 = 9
hg_n_base_list_pse_proba = []
hg_pse_list_num = []
M_threshold = Threshold.choiceThreshold(species, modification, M_threshold)
# -----------------------------PSE ---------------------------------------------------
for base in line.strip()[10:-10]:
hg_n_base_21 += 1
file_inter_pse = open('inter_use/hg_inter_use_pse.txt', 'w')
if base == 'U':
new_line = line.strip()[hg_n_base_21 - 10:hg_n_base_21 + 11]
# ---f nc dnc tnc tenc pnc nd dnd ncp dpcp onehot dbe-----------------
print(*iMRMFeatureEtraction.nc(new_line, 21),
*iMRMFeatureEtraction.dnc(new_line, 21),
*iMRMFeatureEtraction.tnc(new_line, 21),
*iMRMFeatureEtraction.tenc(new_line, 21),
*iMRMFeatureEtraction.pnc(new_line, 21),
*iMRMFeatureEtraction.nd(new_line, 21),
*iMRMFeatureEtraction.dnd(new_line, 21),
*iMRMFeatureEtraction.ncp(new_line, 21),
*iMRMFeatureEtraction.dpcp(new_line, 21),
*iMRMFeatureEtraction.onehot(new_line, 21),
*iMRMFeatureEtraction.dbe(new_line, 21),
file=file_inter_pse)
file_inter_pse.close()
pddtest = pd.read_csv("inter_use/hg_inter_use_pse.txt",
sep=' ',
header=None)
X_test = pddtest.values[:, [
x - 1 for x in hg_importance_dict_sorted_list_num_Pse
]]
prediction = XG_iRNA_hg_Pse.predict_proba(X_test)
if prediction[0][1] > M_threshold:
hg_n_base_list_pse_proba.append(prediction[0][1])
hg_pse_list_num.append(hg_n_base_21 + 1)
if len(hg_pse_list_num) > 0:
hg_pse_list_num.reverse()
hg_n_base_list_pse_proba.reverse()
n = 0
for i in [i - 1 for i in hg_pse_list_num]:
line = list(line)
line[i] = '<font color=#FF3333>U</font>(' + str(
hg_n_base_list_pse_proba[n]) + ')'
n += 1
line = str(line).replace(',', '')
line = line.replace('[', '')
line = line.replace(']', '')
line = line.replace(' ', '')
line = line.replace('\'', '')
line = line.replace('\\r', '')
line_pro = line.replace('fontcolor', 'font color')
line = line_pro
pro_list = re.findall('\(.*?\)', line_pro)
for i in pro_list:
line = line.replace(i, '')
return line, line_pro, hg_pse_list_num
from imutils.video import FileVideoStream
from imutils.video import FPS
from skimage.measure import compare_ssim
import numpy as np
import argparse
import imutils
import time
import cv2
import os
import datetime
from pydarknet import Detector, Image
import Threshold
videoFile="HighlineGate.mp4"
yolodir="yolo"
threshold = Threshold.Threshold(videoFile).get()
# parse the arguments
ap = argparse.ArgumentParser()
ap.add_argument("-y", "--yolo", default=yolodir,
help="base path to YOLO directory")
ap.add_argument("-c", "--confidence", type=float, default=0.5,
help="minimum probability to filter weak detections")
ap.add_argument("-s", "--sample", default=videoFile,
help="minimum probability to filter weak detections")
ap.add_argument("-st", "--thresh", default=threshold,
help="minimum probability to filter weak detections")
args = vars(ap.parse_args())
def sc_XG_iRNA_M5C(sequence, species, modification, M_threshold):
M_threshold = Threshold.choiceThreshold(species, modification, M_threshold)
importance_dict_sorted_list_num_M5C = [
299, 1009, 705, 1542, 1522, 1008, 1764, 4, 1862, 1200, 142, 1385, 1358,
1214, 1842, 27, 693, 1086, 193, 1527, 1904, 1533, 1554, 920, 323, 1879,
16
]
XG_iRNA_sc_M5C = load('model/sc/sc_xgb_M5C_10fold')
# --------------------------------------------------------------
line = sequence # bug
if re.search(r'[^AUCG]', line.strip()):
nota = "Please input the RNA sequence 'AUCG'."
return nota
length = len(line.strip())
if length < 41:
nota = "Please input the RNA sequence at least 51 bases! "
return nota
sc_n_base_41_M5C = 19
sc_n_base_list_M5C_proba = []
sc_M5C_list_num = []
for base in line.strip()[20:-20]:
sc_n_base_41_M5C += 1
file_inter_M5C = open('inter_use/sc_inter_use_M5C.txt', 'w')
if base == 'C':
new_line = line.strip()[sc_n_base_41_M5C - 20:sc_n_base_41_M5C +
21]
# ---f nc dnc tnc tenc pnc nd dnd ncp dpcp onehot dbe-----------------
print(*iMRMFeatureEtraction.nc(new_line, 41),
*iMRMFeatureEtraction.dnc(new_line, 41),
*iMRMFeatureEtraction.tnc(new_line, 41),
*iMRMFeatureEtraction.tenc(new_line, 41),
*iMRMFeatureEtraction.pnc(new_line, 41),
*iMRMFeatureEtraction.nd(new_line, 41),
*iMRMFeatureEtraction.dnd(new_line, 41),
*iMRMFeatureEtraction.ncp(new_line, 41),
*iMRMFeatureEtraction.dpcp(new_line, 41),
*iMRMFeatureEtraction.onehot(new_line, 41),
*iMRMFeatureEtraction.dbe(new_line, 41),
file=file_inter_M5C)
file_inter_M5C.close()
pddtest = pd.read_csv("inter_use/sc_inter_use_M5C.txt",
sep=' ',
header=None)
X_test = pddtest.values[:, [
x - 1 for x in importance_dict_sorted_list_num_M5C
]]
prediction = XG_iRNA_sc_M5C.predict_proba(X_test)
if prediction[0][1] > M_threshold:
sc_n_base_list_M5C_proba.append(prediction[0][1])
sc_M5C_list_num.append(sc_n_base_41_M5C + 1)
if len(sc_M5C_list_num) > 0:
sc_M5C_list_num.reverse()
sc_n_base_list_M5C_proba.reverse()
n = 0
for i in [i - 1 for i in sc_M5C_list_num]:
line = list(line)
line[i] = '<font color=#FFA500>C</font>(' + str(
sc_n_base_list_M5C_proba[n]) + ')'
n += 1
line = str(line).replace(',', '')
line = line.replace('[', '')
line = line.replace(']', '')
line = line.replace(' ', '')
line = line.replace('\'', '')
line = line.replace('\\r', '')
line_pro = line.replace('fontcolor', 'font color')
line = line_pro
pro_list = re.findall('\(.*?\)', line_pro)
for i in pro_list:
line = line.replace(i, '')
return line, line_pro, sc_M5C_list_num
def sc_XG_iRNA_M6A(sequence, species, modification, M_threshold):
M_threshold = Threshold.choiceThreshold(species, modification, M_threshold)
importance_dict_sorted_list_num_M6A = [
1912, 1879, 1905, 1310, 327, 11, 1, 1536, 1309, 53, 18, 1311, 762, 280,
474, 4, 34, 71, 1908, 1555, 1549, 1142, 728, 186, 29, 7, 1554, 43, 46,
1117, 75, 1943, 100, 1559, 872, 126, 135, 650, 58, 1411, 981, 542,
1577, 119, 568, 216, 1441, 552
]
XG_iRNA_sc_M6A = load('model/sc/sc_xgb_M6A_10fold')
# -------------------------------------------------------------
line = sequence
if re.search(r'[^AUCG]', line.strip()):
nota = "Please input the RNA sequence 'AUCG'."
return nota
length = len(line.strip())
if length < 41:
nota = "Please input the RNA sequence at least 51 bases! "
return nota
sc_n_base_41_M6A = 19
sc_n_base_list_M6A_proba = []
sc_M6A_list_num = []
for base in line.strip()[20:-20]:
sc_n_base_41_M6A += 1
file_inter_M6A = open('inter_use/sc_inter_use_M6A.txt', 'w')
if base == 'A':
new_line = line.strip()[sc_n_base_41_M6A - 20:sc_n_base_41_M6A +
21]
# ---f nc dnc tnc tenc pnc nd dnd ncp dpcp onehot dbe-----------------
print(*iMRMFeatureEtraction.nc(new_line, 41),
*iMRMFeatureEtraction.dnc(new_line, 41),
*iMRMFeatureEtraction.tnc(new_line, 41),
*iMRMFeatureEtraction.tenc(new_line, 41),
*iMRMFeatureEtraction.pnc(new_line, 41),
*iMRMFeatureEtraction.nd(new_line, 41),
*iMRMFeatureEtraction.dnd(new_line, 41),
*iMRMFeatureEtraction.ncp(new_line, 41),
*iMRMFeatureEtraction.dpcp(new_line, 41),
*iMRMFeatureEtraction.onehot(new_line, 41),
*iMRMFeatureEtraction.dbe(new_line, 41),
file=file_inter_M6A)
file_inter_M6A.close()
pddtest = pd.read_csv("inter_use/sc_inter_use_M6A.txt",
sep=' ',
header=None)
X_test = pddtest.values[:, [
x - 1 for x in importance_dict_sorted_list_num_M6A
]]
prediction = XG_iRNA_sc_M6A.predict_proba(X_test)
if prediction[0][1] > M_threshold:
sc_n_base_list_M6A_proba.append(prediction[0][1])
sc_M6A_list_num.append(sc_n_base_41_M6A + 1)
if len(sc_M6A_list_num) > 0:
sc_M6A_list_num.reverse()
sc_n_base_list_M6A_proba.reverse()
n = 0
for i in [i - 1 for i in sc_M6A_list_num]:
line = list(line)
line[i] = '<font color=#00FF00>A</font>(' + str(
sc_n_base_list_M6A_proba[n]) + ')'
n += 1
line = str(line).replace(',', '')
line = line.replace('[', '')
line = line.replace(']', '')
line = line.replace(' ', '')
line = line.replace('\'', '')
line = line.replace('\\r', '')
line_pro = line.replace('fontcolor', 'font color')
line = line_pro
pro_list = re.findall('\(.*?\)', line_pro)
for i in pro_list:
line = line.replace(i, '')
return line, line_pro, sc_M6A_list_num
def sc_XG_iRNA_M1A(sequence, species, modification, M_threshold):
# -------------------------------------------------------------
M_threshold = Threshold.choiceThreshold(species, modification, M_threshold)
importance_dict_sorted_list_num_M1A = [1824, 1831, 1500]
XG_iRNA_sc_M1A = load('model/sc/sc_xgb_M1A_10fold')
# --------------------------------------------------------------
line = sequence
# bug
if re.search(r'[^AUCG]', line.strip()):
nota = "Please input the RNA sequence 'AUCG'."
return nota
length = len(line.strip())
if length < 41:
nota = "Please input the RNA sequence at least 51 bases! "
return nota
sc_n_base_41_M1A = 19
sc_n_base_list_M1A_proba = []
sc_M1A_list_num = []
for base in line.strip()[20:-20]:
sc_n_base_41_M1A += 1
file_inter_M1A = open('inter_use/sc_inter_use_M1A.txt', 'w')
if base == 'A':
new_line = line.strip()[sc_n_base_41_M1A - 20:sc_n_base_41_M1A +
21]
# ---f nc dnc tnc tenc pnc nd dnd ncp dpcp onehot dbe-----------------
print(*iMRMFeatureEtraction.nc(new_line, 41),
*iMRMFeatureEtraction.dnc(new_line, 41),
*iMRMFeatureEtraction.tnc(new_line, 41),
*iMRMFeatureEtraction.tenc(new_line, 41),
*iMRMFeatureEtraction.pnc(new_line, 41),
*iMRMFeatureEtraction.nd(new_line, 41),
*iMRMFeatureEtraction.dnd(new_line, 41),
*iMRMFeatureEtraction.ncp(new_line, 41),
*iMRMFeatureEtraction.dpcp(new_line, 41),
*iMRMFeatureEtraction.onehot(new_line, 41),
*iMRMFeatureEtraction.dbe(new_line, 41),
file=file_inter_M1A)
file_inter_M1A.close()
pddtest = pd.read_csv("inter_use/sc_inter_use_M1A.txt",
sep=' ',
header=None)
X_test = pddtest.values[:, [
x - 1 for x in importance_dict_sorted_list_num_M1A
]]
prediction = XG_iRNA_sc_M1A.predict_proba(X_test)
if prediction[0][1] > M_threshold:
sc_n_base_list_M1A_proba.append(prediction[0][1])
sc_M1A_list_num.append(sc_n_base_41_M1A + 1)
if len(sc_M1A_list_num) > 0:
sc_M1A_list_num.reverse()
sc_n_base_list_M1A_proba.reverse()
n = 0
for i in [i - 1 for i in sc_M1A_list_num]:
line = list(line)
line[i] = '<font color=#1E90FF>A</font>(' + str(
sc_n_base_list_M1A_proba[n]) + ')'
n += 1
line = str(line).replace(',', '')
line = line.replace('[', '')
line = line.replace(']', '')
line = line.replace(' ', '')
line = line.replace('\'', '')
line = line.replace('\\r', '')
line_pro = line.replace('fontcolor', 'font color')
line = line_pro
pro_list = re.findall('\(.*?\)', line_pro)
for i in pro_list:
line = line.replace(i, '')
return line, line_pro, sc_M1A_list_num
def sc_XG_iRNA_pse(sequence, species, modification, M_threshold):
M_threshold = Threshold.choiceThreshold(species, modification, M_threshold)
# -------------------------------------------------------------
sc_importance_dict_sorted_list_num_Pse = [
1477, 201, 1748, 1759, 1380, 1784, 10, 1121, 693, 1492, 135, 1377,
1792, 808, 1383, 251, 1800, 1766, 1746, 1, 302, 146, 78, 1488, 34, 242,
76, 230, 12, 1747, 13, 6, 269, 1753, 714, 38, 182, 77, 324, 97, 165,
1788, 664, 82, 893, 194
]
XG_iRNA_sc_Pse = load('model/sc/sc_xgb_Pse_10fold')
# --------------------------------------------------------------
line = sequence
line = line.upper() # bug
if re.search(r'[^AUCG]', line.strip()):
nota = "Please input the RNA sequence 'AUCG'."
return nota
length = len(line.strip())
if length < 31:
nota = "Please input the RNA sequence at least 51 bases! "
return nota
sc_n_base_31 = 14
sc_n_base_list_pse_proba = []
sc_pse_list_num = []
# -----------------------------PSE ---------------------------------------------------
for base in line.strip()[15:-15]:
sc_n_base_31 += 1
file_inter_pse = open('inter_use/sc_inter_use_pse.txt', 'w')
if base == 'U':
new_line = line.strip()[sc_n_base_31 - 15:sc_n_base_31 + 16]
# ---f nc dnc tnc tenc pnc nd dnd ncp dpcp onehot dbe-----------------
print(*iMRMFeatureEtraction.nc(new_line, 31),
*iMRMFeatureEtraction.dnc(new_line, 31),
*iMRMFeatureEtraction.tnc(new_line, 31),
*iMRMFeatureEtraction.tenc(new_line, 31),
*iMRMFeatureEtraction.pnc(new_line, 31),
*iMRMFeatureEtraction.nd(new_line, 31),
*iMRMFeatureEtraction.dnd(new_line, 31),
*iMRMFeatureEtraction.ncp(new_line, 31),
*iMRMFeatureEtraction.dpcp(new_line, 31),
*iMRMFeatureEtraction.onehot(new_line, 31),
*iMRMFeatureEtraction.dbe(new_line, 31),
file=file_inter_pse)
file_inter_pse.close()
pddtest = pd.read_csv("inter_use/sc_inter_use_pse.txt",
sep=' ',
header=None)
X_test = pddtest.values[:, [
x - 1 for x in sc_importance_dict_sorted_list_num_Pse
]]
prediction = XG_iRNA_sc_Pse.predict_proba(X_test)
if prediction[0][1] > M_threshold:
sc_n_base_list_pse_proba.append(prediction[0][1])
sc_pse_list_num.append(sc_n_base_31 + 1)
if len(sc_pse_list_num) > 0:
sc_pse_list_num.reverse()
sc_n_base_list_pse_proba.reverse()
n = 0
for i in [i - 1 for i in sc_pse_list_num]:
line = list(line)
line[i] = '<font color=#FF3333>U</font>(' + str(
sc_n_base_list_pse_proba[n]) + ')'
n += 1
line = str(line).replace(',', '')
line = line.replace('[', '')
line = line.replace(']', '')
line = line.replace(' ', '')
line = line.replace('\'', '')
line = line.replace('\\r', '')
line_pro = line.replace('fontcolor', 'font color')
line = line_pro
pro_list = re.findall('\(.*?\)', line_pro)
for i in pro_list:
line = line.replace(i, '')
return line, line_pro, sc_pse_list_num
def human_XG_iRNA_M5C(sequence, species, modification, M_threshold):
importance_dict_sorted_list_num_M5C = [
16, 144, 1442, 82, 18, 1225, 210, 1401, 1430, 80, 57, 68, 1562, 74, 950
]
XG_iRNA_hg_M5C = load('model/hg/hg_xgb_M5C_10fold')
M_threshold = Threshold.choiceThreshold(species, modification, M_threshold)
# --------------------------------------------------------------
line = sequence
line = line.upper() # bug
if re.search(r'[^AUCG]', line.strip()):
nota = "Please input the RNA sequence 'AUCG'."
return nota
length = len(line.strip())
if length < 41:
nota = "Please input the RNA sequence at least 51 bases! "
return nota
hg_n_base_41_M5C = 19
hg_n_base_list_M5C_proba = []
hg_M5C_list_num = []
for base in line.strip()[20:-20]:
hg_n_base_41_M5C += 1
file_inter_M5C = open('inter_use/hg_inter_use_M5C.txt', 'w')
if base == 'C':
new_line = line.strip()[hg_n_base_41_M5C - 20:hg_n_base_41_M5C +
21]
# ---f nc dnc tnc tenc pnc nd dnd ncp dpcp onehot dbe-----------------
print(*iMRMFeatureEtraction.nc(new_line, 41),
*iMRMFeatureEtraction.dnc(new_line, 41),
*iMRMFeatureEtraction.tnc(new_line, 41),
*iMRMFeatureEtraction.tenc(new_line, 41),
*iMRMFeatureEtraction.pnc(new_line, 41),
*iMRMFeatureEtraction.nd(new_line, 41),
*iMRMFeatureEtraction.dnd(new_line, 41),
*iMRMFeatureEtraction.ncp(new_line, 41),
*iMRMFeatureEtraction.dpcp(new_line, 41),
*iMRMFeatureEtraction.onehot(new_line, 41),
*iMRMFeatureEtraction.dbe(new_line, 41),
file=file_inter_M5C)
file_inter_M5C.close()
pddtest = pd.read_csv("inter_use/hg_inter_use_M5C.txt",
sep=' ',
header=None)
X_test = pddtest.values[:, [
x - 1 for x in importance_dict_sorted_list_num_M5C
]]
prediction = XG_iRNA_hg_M5C.predict_proba(X_test)
if prediction[0][1] > M_threshold:
hg_n_base_list_M5C_proba.append(prediction[0][1])
hg_M5C_list_num.append(hg_n_base_41_M5C + 1)
if len(hg_M5C_list_num) > 0:
hg_M5C_list_num.reverse()
hg_n_base_list_M5C_proba.reverse()
n = 0
for i in [i - 1 for i in hg_M5C_list_num]:
line = list(line)
line[i] = '<font color=#FFA500>C</font>(' + str(
hg_n_base_list_M5C_proba[n]) + ')'
n += 1
line = str(line).replace(',', '')
line = line.replace('[', '')
line = line.replace(']', '')
line = line.replace(' ', '')
line = line.replace('\'', '')
line = line.replace('\\r', '')
line_pro = line.replace('fontcolor', 'font color')
line = line_pro
pro_list = re.findall('\(.*?\)', line_pro)
for i in pro_list:
line = line.replace(i, '')
return line, line_pro, hg_M5C_list_num
def human_XG_iRNA_M6A(sequence, species, modification, M_threshold):
importance_dict_sorted_list_num_M6A = [
1834, 1500, 1831, 1503, 16, 1446, 1489, 1882, 30, 113, 997, 1545, 1857,
1892, 87, 4, 147, 1866, 209, 369, 1395, 1366, 1520, 1401, 1422, 1397,
1398, 279, 1454, 40, 243, 1417, 1498, 42, 81, 377, 1441, 51, 12, 1557,
17, 404, 96, 724
]
XG_iRNA_hg_M6A = load('model/hg/hg_xgb_M6A_10fold')
M_threshold = Threshold.choiceThreshold(species, modification, M_threshold)
# -------------------------------------------------------------
line = sequence
line = line.upper() # bug
if re.search(r'[^AUCG]', line.strip()):
nota = "Please input the RNA sequence 'AUCG'."
return nota
length = len(line.strip())
if length < 41:
nota = "Please input the RNA sequence at least 51 bases! "
return nota
hg_n_base_41_M6A = 19
hg_n_base_list_M6A_proba = []
hg_M6A_list_num = []
for base in line.strip()[20:-20]:
hg_n_base_41_M6A += 1
file_inter_M6A = open('inter_use/hg_inter_use_M6A.txt', 'w')
if base == 'A':
new_line = line.strip()[hg_n_base_41_M6A - 20:hg_n_base_41_M6A +
21]
# ---f nc dnc tnc tenc pnc nd dnd ncp dpcp onehot dbe-----------------
print(*iMRMFeatureEtraction.nc(new_line, 41),
*iMRMFeatureEtraction.dnc(new_line, 41),
*iMRMFeatureEtraction.tnc(new_line, 41),
*iMRMFeatureEtraction.tenc(new_line, 41),
*iMRMFeatureEtraction.pnc(new_line, 41),
*iMRMFeatureEtraction.nd(new_line, 41),
*iMRMFeatureEtraction.dnd(new_line, 41),
*iMRMFeatureEtraction.ncp(new_line, 41),
*iMRMFeatureEtraction.dpcp(new_line, 41),
*iMRMFeatureEtraction.onehot(new_line, 41),
*iMRMFeatureEtraction.dbe(new_line, 41),
file=file_inter_M6A)
file_inter_M6A.close()
pddtest = pd.read_csv("inter_use/hg_inter_use_M6A.txt",
sep=' ',
header=None)
X_test = pddtest.values[:, [
x - 1 for x in importance_dict_sorted_list_num_M6A
]]
prediction = XG_iRNA_hg_M6A.predict_proba(X_test)
if prediction[0][1] > M_threshold:
hg_n_base_list_M6A_proba.append(prediction[0][1])
hg_M6A_list_num.append(hg_n_base_41_M6A + 1)
if len(hg_M6A_list_num) > 0:
hg_M6A_list_num.reverse()
hg_n_base_list_M6A_proba.reverse()
n = 0
for i in [i - 1 for i in hg_M6A_list_num]:
line = list(line)
line[i] = '<font color=#00FF00>A</font>(' + str(
hg_n_base_list_M6A_proba[n]) + ')'
n += 1
line = str(line).replace(',', '')
line = line.replace('[', '')
line = line.replace(']', '')
line = line.replace(' ', '')
line = line.replace('\'', '')
line = line.replace('\\r', '')
line_pro = line.replace('fontcolor', 'font color')
line = line_pro
pro_list = re.findall('\(.*?\)', line_pro)
for i in pro_list:
line = line.replace(i, '')
return line, line_pro, hg_M6A_list_num
def human_XG_iRNA_M1A(sequence, species, modification, M_threshold):
# -------------------------------------------------------------
importance_dict_sorted_list_num_M1A = [
1834, 1831, 16, 1500, 1503, 3, 9, 1, 19, 4, 20, 6, 2, 1378, 1386, 1872,
30, 1520, 1529, 1484, 243, 1002, 1372, 259, 1873, 200, 1817, 312, 1400,
584, 572, 1404, 943, 239, 41, 190, 1415, 34, 1455, 64, 1856
]
XG_iRNA_hg_M1A = load('model/hg/hg_xgb_M1A_10fold')
# --------------------------------------------------------------
M_threshold = Threshold.choiceThreshold(species, modification, M_threshold)
line = sequence # bug
hg_n_base_41_M1A = 19
hg_n_base_list_M1A_proba = []
hg_M1A_list_num = []
for base in line.strip()[20:-20]:
hg_n_base_41_M1A += 1
file_inter_M1A = open('inter_use/hg_inter_use_M1A.txt', 'w')
if base == 'A':
new_line = line.strip()[hg_n_base_41_M1A - 20:hg_n_base_41_M1A +
21]
# ---f nc dnc tnc tenc pnc nd dnd ncp dpcp onehot dbe-----------------
print(*iMRMFeatureEtraction.nc(new_line, 41),
*iMRMFeatureEtraction.dnc(new_line, 41),
*iMRMFeatureEtraction.tnc(new_line, 41),
*iMRMFeatureEtraction.tenc(new_line, 41),
*iMRMFeatureEtraction.pnc(new_line, 41),
*iMRMFeatureEtraction.nd(new_line, 41),
*iMRMFeatureEtraction.dnd(new_line, 41),
*iMRMFeatureEtraction.ncp(new_line, 41),
*iMRMFeatureEtraction.dpcp(new_line, 41),
*iMRMFeatureEtraction.onehot(new_line, 41),
*iMRMFeatureEtraction.dbe(new_line, 41),
file=file_inter_M1A)
file_inter_M1A.close()
pddtest = pd.read_csv("inter_use/hg_inter_use_M1A.txt",
sep=' ',
header=None)
X_test = pddtest.values[:, [
x - 1 for x in importance_dict_sorted_list_num_M1A
]]
prediction = XG_iRNA_hg_M1A.predict_proba(X_test)
if prediction[0][1] > M_threshold:
hg_n_base_list_M1A_proba.append(prediction[0][1])
hg_M1A_list_num.append(hg_n_base_41_M1A + 1)
if len(hg_M1A_list_num) > 0:
hg_M1A_list_num.reverse()
hg_n_base_list_M1A_proba.reverse()
n = 0
for i in [i - 1 for i in hg_M1A_list_num]:
line = list(line)
line[i] = '<font color=#1E90FF>A</font>(' + str(
hg_n_base_list_M1A_proba[n]) + ')'
n += 1
line = str(line).replace(',', '')
line = line.replace('[', '')
line = line.replace(']', '')
line = line.replace(' ', '')
line = line.replace('\'', '')
line = line.replace('\\r', '')
line_pro = line.replace('fontcolor', 'font color')
line = line_pro
pro_list = re.findall('\(.*?\)', line_pro)
for i in pro_list:
line = line.replace(i, '')
return line, line_pro, hg_M1A_list_num
def process_image(mtx, dist, image):
# Define an image to draw on
draw_image = np.copy(image)
# **LANE LINE DETECTION**
# Undistort the image
dist = cv2.undistort(image, mtx, dist, None, mtx)
# Apply thresholding to the image
combined_binary = Threshold.white_yellow_threshold(dist)
# Mask lane line area
masked_img = FindLaneLineUtils.mask_lane_line_area(combined_binary)
# Warp the image to an overhead view
warped_img, src, dest = FindLaneLineUtils.warp_image(masked_img)
# Identify the lane line pixels
combined = FindLaneLineUtils.get_lane_line_pixels(warped_img)
# Get polynomials for lane lines
left_fit, right_fit = FindLaneLineUtils.get_polynomials(combined)
# **VEHICLE DETECTION**
# Define xy overlap and use slide_windows function to develop search windows for vehicles
xy_overlap = (0.8, 0.8)
windows_1 = slide_window(image,
x_start_stop=[None, None],
y_start_stop=[440, 620],
xy_window=(128, 128),
xy_overlap=xy_overlap)
windows_2 = slide_window(image,
x_start_stop=[None, None],
y_start_stop=[400, 496],
xy_window=(96, 96),
xy_overlap=xy_overlap)
windows_3 = slide_window(image,
x_start_stop=[None, None],
y_start_stop=[400, 528],
xy_window=(64, 64),
xy_overlap=xy_overlap)
windows = windows_1 + windows_2 + windows_3
# Alter images read since images are jpg, where training data is png
sized_image = image.astype(np.float32) / 255
# Use hot_windows function to "turn on" windows are assigned as vehicle by the classifier
hot_windows = search_windows(sized_image,
windows,
svc,
X_scaler,
color_space=color_space,
spatial_size=spatial_size,
hist_bins=hist_bins,
orient=orient,
pix_per_cell=pix_per_cell,
cell_per_block=cell_per_block,
hog_channel=hog_channel,
spatial_feat=spatial_feat,
hist_feat=hist_feat,
hog_feat=hog_feat)
# Intialize all_hot_windows with this frame's hot windows
all_hot_windows = hot_windows
# Define window image, draw_boxes draws all_hot_windows onto the draw image
window_img = draw_boxes(draw_image,
all_hot_windows,
color=(0, 0, 255),
thick=6)
# Add heat to each box in box list
heat = np.zeros_like(image[:, :, 0]).astype(np.float)
heat = add_heat(heat, all_hot_windows)
# Apply a threshold to the heatmap
threshold = (np.array(all_hot_windows).shape[0]) // 10 + 1
# threshold = 1
heat = apply_threshold(heat, threshold)
# Visualize the heatmap when displaying
heatmap = np.clip(heat, 0, 255)
# mpimg.imsave((os.path.join("output_images/heat_images/", filename)), heatmap)
# Find final boxes from heatmap using label function
# save the boxes drawn as an array
labels = label(heat)
output_boxes = get_vehicle_boxes(np.copy(image), labels)
# **MONITOR OVER LAST 10 FRAMES**
# Update results
results.update(output_boxes, left_fit, right_fit)
# Get the average over the last 10 frames
avg_boxes = results.get_boxes()
avg_left_fit = results.get_poly_left()
avg_right_fit = results.get_poly_right()
# **COMBINED DETECTION & DRAW**
# Draw the lane area on a blank image
poly_img = FindLaneLineUtils.plot_polynomial_overlay(
combined, avg_left_fit, avg_right_fit)
# Un-warp the image back to it's normal view with the lane line area overlain
lane_line_overlay_img = FindLaneLineUtils.overlay_with_inverse_transform(
draw_image, poly_img, src, dest)
# Add the vehicle detection boxes to to the lane line overlay image
box_image = draw_labeled_bboxes(np.copy(image), avg_boxes)
result = cv2.addWeighted(box_image, 0.5, lane_line_overlay_img, 0.5, 0)
return result
import Sobel_edge_enhancement as enhance
import Threshold as detect
import time
#Create a grayscale image
grayImage = gray.GrayScale()
#Apply a Gaussian Blur filter
blurImage = blur.GaussianBlur(grayImage)
#Perform Edge Enhancement
start = time.time()
enhancedImage = enhance.sobel_edge_detection(blurImage)
end = time.time()
print(f"Edge enhancement completed in {round(end-start,2)} seconds.")
#Perform Sobel Edge Detection
start = time.time()
finalImage = detect.Threshold()
end = time.time()
print(f"Final edge detection completed in {round(end-start,2)} seconds.")
'''
===============================================================================
ACADEMIC INTEGRITY STATEMENT
I have not used source code obtained from any other unauthorized
source, either modified or unmodified. Neither have I provided
access to my code to another. The project I am submitting
is my own original work.
===============================================================================
'''
def mm_XG_iRNA_pse(sequence, species, modification, M_threshold):
M_threshold = Threshold.choiceThreshold(species, modification, M_threshold)
# -------------------------------------------------------------
mm_importance_dict_sorted_list_num_Pse = [
1441, 1442, 85, 1692, 55, 190, 1678, 101, 25, 221, 122, 128, 60, 692,
93, 16, 255, 173, 1696, 12, 35, 701, 1715, 67, 183, 150, 808, 147,
1431, 43, 1463, 5, 1378, 71, 44, 1377, 1385, 1440, 1446, 1373, 278,
1649, 1697, 1368, 49, 662, 1457, 327, 1384, 767
]
XG_iRNA_mm_Pse = load('model/mm/mm_xgb_Pse_10fold')
# --------------------------------------------------------------
line = sequence # bug
if re.search(r'[^AUCG]', line.strip()):
nota = "Please input the RNA sequence 'AUCG'."
return nota
length = len(line.strip())
if length < 21:
nota = "Please input the RNA sequence at least 51 bases! "
return nota
mm_n_base_21 = 9
mm_n_base_list_pse_proba = []
mm_pse_list_num = []
# -----------------------------PSE ---------------------------------------------------
for base in line.strip()[10:-10]:
mm_n_base_21 += 1
file_inter_pse = open('inter_use/mm_inter_use_pse.txt', 'w')
if base == 'U':
new_line = line.strip()[mm_n_base_21 - 10:mm_n_base_21 + 11]
# ---f nc dnc tnc tenc pnc nd dnd ncp dpcp onehot dbe-----------------
print(*iMRMFeatureEtraction.nc(new_line, 21),
*iMRMFeatureEtraction.dnc(new_line, 21),
*iMRMFeatureEtraction.tnc(new_line, 21),
*iMRMFeatureEtraction.tenc(new_line, 21),
*iMRMFeatureEtraction.pnc(new_line, 21),
*iMRMFeatureEtraction.nd(new_line, 21),
*iMRMFeatureEtraction.dnd(new_line, 21),
*iMRMFeatureEtraction.ncp(new_line, 21),
*iMRMFeatureEtraction.dpcp(new_line, 21),
*iMRMFeatureEtraction.onehot(new_line, 21),
*iMRMFeatureEtraction.dbe(new_line, 21),
file=file_inter_pse)
file_inter_pse.close()
pddtest = pd.read_csv("inter_use/mm_inter_use_pse.txt",
sep=' ',
header=None)
X_test = pddtest.values[:, [
x - 1 for x in mm_importance_dict_sorted_list_num_Pse
]]
prediction = XG_iRNA_mm_Pse.predict_proba(X_test)
if prediction[0][1] > M_threshold:
mm_n_base_list_pse_proba.append(prediction[0][1])
mm_pse_list_num.append(mm_n_base_21 + 1)
if len(mm_pse_list_num) > 0:
mm_pse_list_num.reverse()
mm_n_base_list_pse_proba.reverse()
n = 0
for i in [i - 1 for i in mm_pse_list_num]:
line = list(line)
line[i] = '<font color=#FF3333>U</font>(' + str(
mm_n_base_list_pse_proba[n]) + ')'
n += 1
line = str(line).replace(',', '')
line = line.replace('[', '')
line = line.replace(']', '')
line = line.replace(' ', '')
line = line.replace('\'', '')
line = line.replace('\\r', '')
line_pro = line.replace('fontcolor', 'font color')
line = line_pro
pro_list = re.findall('\(.*?\)', line_pro)
for i in pro_list:
line = line.replace(i, '')
return line, line_pro, mm_pse_list_num
def human_XG_iRNA_AI(sequence, species, modification, M_threshold):
hg_importance_dict_sorted_list_num_AI = [
1899, 1895, 5, 1, 69, 79, 252, 1892, 19, 1540, 1390, 1441, 1391, 52,
299, 1931, 1406, 25, 572, 1547, 83, 1983, 303, 1367, 72, 1414, 23, 250,
61, 110, 1584, 24, 401, 1402, 1449, 1903, 1437, 341, 214, 1894, 1833,
36, 1574, 739, 1382, 44, 1404, 1526
]
XG_iRNA_hg_AI = load('model/hg/hg_xgb_AI_10fold')
# --------------------------------------------------------------
line = sequence
hg_n_base_51 = 24
hg_n_base_list_AI_proba = []
hg_AI_list_num = []
M_threshold = Threshold.choiceThreshold(species, modification, M_threshold)
# -----------------------------AI ---------------------------------------------------
for base in line.strip()[25:-25]:
hg_n_base_51 += 1
file_inter_AI = open('inter_use/hg_inter_use_AI.txt', 'w')
if base == 'A':
new_line = line.strip()[hg_n_base_51 - 25:hg_n_base_51 + 26]
# ---f nc dnc tnc tenc pnc nd dnd ncp dpcp onehot dbe-----------------
print(*iMRMFeatureEtraction.nc(new_line, 51),
*iMRMFeatureEtraction.dnc(new_line, 51),
*iMRMFeatureEtraction.tnc(new_line, 51),
*iMRMFeatureEtraction.tenc(new_line, 51),
*iMRMFeatureEtraction.pnc(new_line, 51),
*iMRMFeatureEtraction.nd(new_line, 51),
*iMRMFeatureEtraction.dnd(new_line, 51),
*iMRMFeatureEtraction.ncp(new_line, 51),
*iMRMFeatureEtraction.dpcp(new_line, 51),
*iMRMFeatureEtraction.onehot(new_line, 51),
*iMRMFeatureEtraction.dbe(new_line, 51),
file=file_inter_AI)
file_inter_AI.close()
pddtest = pd.read_csv("inter_use/hg_inter_use_AI.txt",
sep=' ',
header=None)
X_test = pddtest.values[:, [
x - 1 for x in hg_importance_dict_sorted_list_num_AI
]]
prediction = XG_iRNA_hg_AI.predict_proba(X_test)
if prediction[0][1] > M_threshold:
hg_n_base_list_AI_proba.append(prediction[0][1])
hg_AI_list_num.append(hg_n_base_51 + 1)
if len(hg_AI_list_num) > 0:
hg_AI_list_num.reverse()
hg_n_base_list_AI_proba.reverse()
n = 0
for i in [i - 1 for i in hg_AI_list_num]:
line = list(line)
line[i] = '<font color=#9400D3>A</font>(' + str(
hg_n_base_list_AI_proba[n]) + ')'
n += 1
line = str(line).replace(',', '')
line = line.replace('[', '')
line = line.replace(']', '')
line = line.replace(' ', '')
line = line.replace('\'', '')
line = line.replace('\\r', '')
line_pro = line.replace('fontcolor', 'font color')
line = line_pro
pro_list = re.findall('\(.*?\)', line_pro)
for i in pro_list:
line = line.replace(i, '')
return line, line_pro, hg_AI_list_num
import numpy as np
import Threshold as th
test = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])
thresholdClass = th.Thresholds(test)
print(thresholdClass.allLabelings())
goodInput = False
while goodInput is False:
choice = int(input("Enter function choice: "))
if choice >= 0 and choice < len(test):
goodInput = True
else:
print("Please enter a number between 0 and the length of the array")
thresholdClass.setBestFunction(choice)
bestThreshold = thresholdClass.getBestFunction()
print(bestThreshold(test))
print(bestThreshold(4))
