@@ -1,6 +1,6 @@

 Package: crisprScore

-Version: 1.1.1

-Date: 2022-04-06

+Version: 1.1.2

+Date: 2022-06-22

 Title: On-Target and Off-Target Scoring Algorithms for CRISPR gRNAs

 Authors@R: c(

     person("Jean-Philippe", "Fortin", email = "[email protected]", role = c("aut", "cre", "cph")),

@@ -37,7 +37,6 @@ importFrom(reticulate,import_from_path)

 importFrom(reticulate,np_array)

 importFrom(reticulate,py_suppress_warnings)

 importFrom(reticulate,r_to_py)

-importFrom(reticulate,source_python)

 importFrom(stats,complete.cases)

 importFrom(stats,predict)

 importFrom(stats,quantile)

@@ -51,41 +51,45 @@ getAzimuthScores <- function(sequences, fork=FALSE){

         stop("Positions 26 and 27 of the sequences must be G",

              " nucleotides (canonical PAM sequences required).")

     }

-    results <- basiliskRun(env=env_azimuth,

-                           shared=FALSE,

-                           fork=fork,

-                           fun=.azimuth_python, 

-                           sequences=sequences)

-    return(results)

-}

-

-

-#' @importFrom reticulate source_python

-#' @importFrom reticulate np_array

-#' @importFrom reticulate import_from_path

-.azimuth_python <- function(sequences){

-

-    dir <- system.file("python",

-                       "azimuth",

-                       package="crisprScore",

-                       mustWork=TRUE)

-    azimuth <- import_from_path("getAzimuth", dir)

     df <- data.frame(sequence=sequences,

                      score=NA_real_,

                      stringsAsFactors=FALSE)

     good <- !grepl("N", sequences)

     sequences.valid <- sequences[good]

-    ns <- length(sequences.valid)

-    if (ns>0){

-        if (ns==1){

-            sequences.valid <- rep(sequences.valid,2)

-            scores <- azimuth$getAzimuth(np_array(sequences.valid))

+

+    #Saving to disk:

+    dir <- tempdir()

+    inputfile  <- file.path(dir, "input.txt")

+    outputfile <- file.path(dir, "output.txt")

+

+      # Ready to get the scores

+    env <- basilisk:::.obtainEnvironmentPath(env_azimuth)

+    basilisk.utils::activateEnvironment(env)

+    programFile <- system.file("python",

+                               "azimuth/getAzimuth.py",

+                               package="crisprScore",

+                               mustWork=TRUE)

+    cmd <- paste0("python ",

+                  programFile, " ",

+                  inputfile, " ",

+                  outputfile)

+

+    if (sum(good)>0){

+        if (sum(good)==1){

+            sequences.valid <- rep(sequences.valid, 2)

+        }

+        .dumpToFile(sequences.valid, inputfile)

+        system(cmd)

+        scores <- read.table(outputfile)[,1]

+        if (sum(good)==1){

             scores <- scores[1]

-        } else {

-            scores <- azimuth$getAzimuth(np_array(sequences.valid)) 

         }

         df$score[good] <- scores

     }

+

     return(df)

 }

+

+

+

@@ -59,35 +59,49 @@ getDeepCpf1Scores <- function(sequences,

             }, FUN.VALUE="character")

         }

     }

-    results <- basiliskRun(env=env_deepcpf1,

-                           shared=FALSE,

-                           fork=fork,

-                           fun=.deepcpf1_python, 

-                           sequences=sequences)

-    return(results)

-}

-

-#' @importFrom reticulate import_from_path

-#' @importFrom reticulate np_array

-#' @importFrom reticulate py_suppress_warnings

-.deepcpf1_python <- function(sequences){

-

-    dir <- system.file("python",

-                       "deepcpf1",

-                       package="crisprScore",

-                       mustWork=TRUE)

-    deepcpf1 <- import_from_path("getDeepCpf1", path=dir)

-    

+  

+    #Output data.frame

     df <- data.frame(sequence=sequences,

                      score=NA_real_,

                      stringsAsFactors=FALSE)

     good <- !grepl("N", sequences)

     sequences.valid <- sequences[good]

-    if (length(sequences.valid)>0){

-        sequences_array <- np_array(sequences.valid)

-        scores <- py_suppress_warnings(deepcpf1$getDeepCpf1(sequences_array))

+

+    #Saving to disk:

+    dir <- tempdir()

+    inputfile  <- file.path(dir, "input.txt")

+    outputfile <- file.path(dir, "output.txt")

+   

+    # Ready to get the scores

+    env <- basilisk:::.obtainEnvironmentPath(env_deepcpf1)

+    basilisk.utils::activateEnvironment(env)

+    programFile <- system.file("python",

+                               "deepcpf1/getDeepCpf1.py",

+                               package="crisprScore",

+                               mustWork=TRUE)

+    cmd <- paste0("python ",

+                  programFile, " ",

+                  inputfile, " ",

+                  outputfile)

+    if (sum(good)>0){

+        .dumpToFile(sequences.valid, inputfile)

+        system(cmd)

+        scores <- read.table(outputfile)[,1]

         scores <- scores/100

         df$score[good] <- scores

-    } 

+    }

+

     return(df)

 }

+

+.dumpToFile <- function(sequences, file){

+    write.table(sequences,

+              file=file,

+              quote=FALSE,

+              col.names=FALSE,

+              row.names=FALSE)

+}

+

+

+

+

@@ -1,7 +1,13 @@

+#sys.argv[1] should be the path of the file with input sequences

+#sys.argv[2] should be the path of the file where to save sequence

 import sys

 import azimuth.model_comparison

 import numpy as np 

 def getAzimuth(sequences):

 	predictions = azimuth.model_comparison.predict(sequences, None, None)

-	return predictions

\ No newline at end of file

+	return predictions

+

+sequences = np.loadtxt(sys.argv[1], dtype="U34", ndmin=1)

+scores = getAzimuth(sequences)

+np.savetxt(sys.argv[2], scores)

@@ -1,3 +1,5 @@

+#sys.argv[1] should be the path of the file with input sequences

+#sys.argv[2] should be the path of the file where to save sequence

 import os

 import numpy as np

 import warnings

@@ -8,4 +10,8 @@ os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'

 def getDeepCpf1(sequences):	

 	results = deepcpf1(sequences)

-	return results

\ No newline at end of file

+	return results

+

+sequences = np.loadtxt(sys.argv[1], dtype="U34", ndmin=1)

+scores = getDeepCpf1(sequences)

+np.savetxt(sys.argv[2], scores)

\ No newline at end of file

@@ -58,6 +58,7 @@ supported at the moment.

 \details{

 \code{tss_df} details:

 This must be a \code{data.frame} that contains the following columns:

+* tss_id: string specifying name of the TSS.

 * gene_symbol: string specifying sHGNC/HUGO gene identifier.

 * promoter: string specifying promoter ID (e.g. "P1" or "P2").

 * transcripts: Ensembl transcript identifier.