updated documentation for softmask support

tutorials/tutorial_predict_and_plot.Rmd

@@ -102,6 +102,7 @@ python ../tensorflow1_version/run_deploy_shape_deepCregr.py --input example_regi
   --name_tag predict \  # name tag to add to ouput files
   --model ./model_deepCregr_5kb_GM12878_primary/model \  # trained deepC model downloaded and extracted
   --genome ./hg19_chr17_fasta_for_test/hg19_chr17.fa \  # link to whole genome or chromosome wise fasta file (needs a fasta index) or test chr17 fasta file dowloaded and extracted
+  --use_softmasked=False \  #Specify if to include base pairs soft masked in the fasta file (lower case) default=False
   --bp_context 1005000 \  # bp context (1 Mb + bin.size)
   --add_window 500000 \  # how much bp to add to either side of the specified window
   --num_classes 201 \  # The number of classes corresponds to the number of outputs (output bins of the vertical pole) (201 for 5kb models; 101 for 10kb models)
@@ -114,6 +115,7 @@ python ../tensorflow1_version/run_deploy_shape_deepCregr.py --input example_regi
   --name_tag predict \
   --model ./model_deepCregr_5kb_GM12878_primary/model \
   --genome ./hg19_chr17_fasta_for_test/hg19_chr17.fa  \
+  --use_softmasked=False  \
   --bp_context 1005000 \
   --add_window 500000 \
   --num_classes 201 \

tutorials/tutorial_predict_and_plot.html

@@ -420,6 +420,7 @@ <h3>Running the deepC prediction</h3>
   --name_tag predict \  # name tag to add to ouput files
   --model ./model_deepCregr_5kb_GM12878_primary/model \  # trained deepC model downloaded and extracted
   --genome ./hg19_chr17_fasta_for_test/hg19_chr17.fa \  # link to whole genome or chromosome wise fasta file (needs a fasta index) or test chr17 fasta file dowloaded and extracted
+  --use_softmasked=False \  #Specify if to include base pairs soft masked in the fasta file (lower case) default=False
   --bp_context 1005000 \  # bp context (1 Mb + bin.size)
   --add_window 500000 \  # how much bp to add to either side of the specified window
   --num_classes 201 \  # The number of classes corresponds to the number of outputs (output bins of the vertical pole) (201 for 5kb models; 101 for 10kb models)
@@ -432,6 +433,7 @@ <h3>Running the deepC prediction</h3>
   --name_tag predict \
   --model ./model_deepCregr_5kb_GM12878_primary/model \
   --genome ./hg19_chr17_fasta_for_test/hg19_chr17.fa  \
+  --use_softmasked=False  \
   --bp_context 1005000 \
   --add_window 500000 \
   --num_classes 201 \
@@ -611,7 +613,7 @@ <h3>Adding the source HiC data</h3>
 # make sure the start and end position match to the binning of the Hi-C data
 # use custom floor/ceiling funcitons for your bin size
 binned.genome &lt;- getBinnedChrom(chr=&quot;chr17&quot;, start=custom_floor(limit1-window.size, bin.size), end=custom_ceil(limit2+window.size, bin.size), window=window.size, step=bin.size)</code></pre>
-<pre><code>## [1] &quot;bedtools makewindows -b tempfile_for_bedtools_call_2020-07-31_09:55:04_0.327818183694035.bed  -w 1005000  -s 5000 &gt;tempfile_from_bedtools_call_2020-07-31_09:55:04_0.327818183694035.bed&quot;</code></pre>
+<pre><code>## [1] &quot;bedtools makewindows -b tempfile_for_bedtools_call_2020-09-04_12:02:11_0.325754458550364.bed  -w 1005000  -s 5000 &gt;tempfile_from_bedtools_call_2020-09-04_12:02:11_0.325754458550364.bed&quot;</code></pre>
 <pre class="r"><code>binned.genome &lt;- as.tibble(binned.genome)
 names(binned.genome) &lt;- c(&quot;chr&quot;, &quot;start&quot;, &quot;end&quot;)
 

tutorials/tutorial_train_a_model.md

@@ -74,6 +74,7 @@ seed_file='./saved_conv_weights_dhw_5layer_1k_pool.npz' #trained filters phase I
 shuffle=True
 store_dtype='bool'  # how to store the sequence
 whg_fasta='./hg19.fa'  # link to whole genome fasta file for retrieving the sequences has to be indexed
+use_softmasked=False  # specify if to use soft masked bases from the fasta file (lowercase). Default=False
 
 # if multiple GPUs present select a single one to run training on
 # and not block the remaining
@@ -111,6 +112,7 @@ python ${SCRIPT_PATH}/run_training_deepCregr.py \
         --shuffle=${shuffle} \
         --store_dtype ${store_dtype} \
         --whg_fasta ${whg_fasta} \
+        --use_softmasked=${use_softmasked} \
         --gpu ${GPU}
 
 ```