medchaabane / deepram Goto Github PK

Hi
I am running the deepRAM as follows
python deepRAM.py --train_data CLIP_train.gz --test_data CLIP_test.gz --data_type RNA --train True --evaluate_performance True --model_path DeepBind.pkl --out_file prediction.txt --Embedding False --Conv True --RNN False --conv_layers 1

and my training and testing files are :
Clip_test.gz
Clip_train.gz

I get the following error when I run it:
python deepRAM.py --train_data Clip_train.gz --test_data Clip_test.gz --data_type RNA --train True --evaluate_performance True --model_path DeepBind.pkl --out_file prediction.txt --Embedding False --Conv True --RNN False --conv_layers 2
cuda:0
False
Load Data
Traceback (most recent call last):
File "deepRAM.py", line 982, in
main()
File "deepRAM.py", line 980, in main
run_deepRAM(args)
File "deepRAM.py", line 896, in run_deepRAM
Load_Data(train_data,test_data)
File "deepRAM.py", line 473, in Load_Data
train1,valid1,train2,valid2,train3,valid3,alldataset,sequences=chipseq.openFile()
File "deepRAM.py", line 318, in openFile
train_dataset.append([seqtopad(row[0],self.motiflen),[int(row[1])]])
File "deepRAM.py", line 36, in seqtopad
if i-motlen+1<len(sequence) and sequence[i-motlen+1]=='N' or ilen(sequence)+motlen-2:
IndexError: string index out of range

I am not sure if the sequence length should be greater than or less than a specific length. Thanks in advance.

Hello, I am learning from DeepRAM for annotation or find motif.
I think this programs very helpful for me.

But, when I try to train as follow your command, i got some syntax error in deepRAM.py

python deepRAM.py --train_data CLIP_train.gz --test_data CLIP_test.gz --data_type RNA --train True --evaluate_performance True --model_path DeepBind.pkl --out_file prediction.txt --Embedding False --Conv True --RNN False --conv_layers 1

File "deepRAM.py", line 276
x=x @self.wHidden
^
SyntaxError: invalid syntax

Please help me. thank you

matplotlib (3.2.1)
I execute the following:
python deepRAM.py --test_data CLIP_test.gz --data_type RNA --predict_only True --model_path DeepBind.pkl --motif True --motif_dir motifs --tomtom_dir meme-5.1.1/src/tomtom --out_file prediction.txt --Embedding False --Conv True --RNN False --conv_layers 1
cuda:0
False
Load Data
Predicting sequence specificities
AUC on test data = 0.7085750683239377
torch.Size([10000, 16, 78])
(16, 4, 24)
plot motif fig motifs
Filter 0
plot logo
Filter 1
plot logo
Filter 2
plot logo
Filter 3
plot logo
Filter 4
plot logo
Filter 5
plot logo
Filter 6
plot logo
Filter 7
plot logo
Filter 8
plot logo
Filter 9
plot logo
Filter 10
plot logo
Filter 11
plot logo
Filter 12
plot logo
Filter 13
plot logo
Filter 14
plot logo
Filter 15
plot logo
The output directory 'motifs/tomtom' already exists.
Its contents will be overwritten.
Processing query 1 out of 16
Estimating pi_0 from all 102 observed p-values.
Estimating pi_0.
Estimated pi_0=0.980392
Processing query 2 out of 16
Estimating pi_0 from all 102 observed p-values.
Estimating pi_0.
Estimated pi_0=0.980392
Processing query 3 out of 16
Estimating pi_0 from all 102 observed p-values.
Estimating pi_0.
Estimated pi_0=1
Processing query 4 out of 16
Estimating pi_0 from all 102 observed p-values.
Estimating pi_0.
Estimated pi_0=0.827383
Processing query 5 out of 16
Estimating pi_0 from all 102 observed p-values.
Estimating pi_0.
Estimated pi_0=0.86333
Processing query 6 out of 16
Estimating pi_0 from all 102 observed p-values.
Estimating pi_0.
Estimated pi_0=0.980392
Processing query 7 out of 16
Estimating pi_0 from all 102 observed p-values.
Estimating pi_0.
Estimated pi_0=0.975416
Processing query 8 out of 16
Estimating pi_0 from all 102 observed p-values.
Estimating pi_0.
Estimated pi_0=0.905859
Processing query 9 out of 16
Estimating pi_0 from all 102 observed p-values.
Estimating pi_0.
Estimated pi_0=0.980392
Processing query 10 out of 16
Estimating pi_0 from all 102 observed p-values.
Estimating pi_0.
Estimated pi_0=1
Processing query 11 out of 16
Estimating pi_0 from all 102 observed p-values.
Estimating pi_0.
Estimated pi_0=0.878533
Processing query 12 out of 16
Estimating pi_0 from all 102 observed p-values.
Estimating pi_0.
Estimated pi_0=0.94579
Processing query 13 out of 16
Estimating pi_0 from all 102 observed p-values.
Estimating pi_0.
Estimated pi_0=0.995172
Processing query 14 out of 16
Estimating pi_0 from all 102 observed p-values.
Estimating pi_0.
Estimated pi_0=0.980392
Processing query 15 out of 16
Estimating pi_0 from all 102 observed p-values.
Estimating pi_0.
Estimated pi_0=0.975365
Processing query 16 out of 16
Estimating pi_0 from all 102 observed p-values.
Estimating pi_0.
Estimated pi_0=0.980392
motifs/tomtom/tomtom.txt motifs_database/Ray2013_rbp_RNA.meme
Traceback (most recent call last):
File "deepRAM.py", line 982, in
main()
File "deepRAM.py", line 980, in main
run_deepRAM(args)
File "deepRAM.py", line 914, in run_deepRAM
Test_Motifs()
File "deepRAM.py", line 755, in Test_Motifs
detect_motifs(model,seqRNA , data, motif_dir)
File "deepRAM.py", line 774, in detect_motifs
get_motif(layer1_para, filter_outs, test_seqs, dir1 = output_dir,embd=embedding,data=data_type,kmer=kmer_len,s=stride,tomtom=tomtom_dir)
File "/scratch/deepRAM/extract_motifs.py", line 698, in get_motif
get_motif_fig(filter_weights, filter_outs, out_dir, testing, sample_i)
File "/scratch/deepRAM/extract_motifs.py", line 635, in get_motif_fig
fmean, fstd = plot_score_density(np.ravel(filter_outs[:,:, f]), '%s/filter%d_dens.pdf' % (out_dir,f))
File "/scratch/deepRAM/extract_motifs.py", line 545, in plot_score_density
sns.distplot(f_scores, kde=False)
File "/usr/lib/python3/dist-packages/seaborn/distributions.py", line 214, in distplot
color=hist_color, **hist_kws)
File "/scratch/.local/lib/python3.6/site-packages/matplotlib/init.py", line 1565, in inner
return func(ax, *map(sanitize_sequence, args), **kwargs)
File "/scratch/.local/lib/python3.6/site-packages/matplotlib/axes/_axes.py", line 6808, in hist
p.update(kwargs)
File "/scratch/.local/lib/python3.6/site-packages/matplotlib/artist.py", line 1006, in update
ret = [_update_property(self, k, v) for k, v in props.items()]
File "/scratch/.local/lib/python3.6/site-packages/matplotlib/artist.py", line 1006, in
ret = [_update_property(self, k, v) for k, v in props.items()]
File "/scratch/.local/lib/python3.6/site-packages/matplotlib/artist.py", line 1002, in _update_property
.format(type(self).name, k))
AttributeError: 'Rectangle' object has no property 'normed'

Thanks for the nice work. We are trying to make your work into an ML problem set for our class.

I have a question about one-hot encoding code here

It seems that the one-hot encoding code set the first few and last few sequences to 0.25. And the length of the sequence that is set to 0.25 is equal to motiflen, I wonder what is the reason for that. I also read the paper, but did not see an explanation for this choice. Is this something standard to do, and where can I read more about this?