Merge pull request #831 from COMBINE-lab/develop

Merge develop -> master

CMakeLists.txt

@@ -1,4 +1,4 @@
-cmake_minimum_required(VERSION 3.12)
+cmake_minimum_required(VERSION 3.15)
 
 if(DEFINED ENV{CC})
   set(CC $ENV{CC})
@@ -662,7 +662,7 @@ tar -xzvf v2021.5.tar.gz
 SOURCE_DIR ${TBB_SOURCE_DIR}
 INSTALL_DIR ${TBB_INSTALL_DIR}
 PATCH_COMMAND "${TBB_PATCH_STEP}"
-CMAKE_ARGS -DCMAKE_CXX_FLAGS=${TBB_CXXFLAGS} -DCMAKE_INSTALL_PREFIX=<INSTALL_DIR> -DTBB_TEST=OFF -DTBB_EXAMPLES=OFF -DCMAKE_BUILD_TYPE=Release -DCMAKE_CXX_STANDARD=${CMAKE_CXX_STANDARD}
+CMAKE_ARGS -DCMAKE_CXX_FLAGS=${TBB_CXXFLAGS} -DCMAKE_INSTALL_PREFIX=<INSTALL_DIR> -DTBB_TEST=OFF -DTBB_EXAMPLES=OFF -DCMAKE_BUILD_TYPE=Release -DCMAKE_CXX_STANDARD=${CMAKE_CXX_STANDARD} -DCMAKE_CXX_COMPILER=${CMAKE_CXX_COMPILER} -DCMAKE_C_COMPILER=${CMAKE_C_COMPILER}
 BUILD_IN_SOURCE TRUE
 )
 
@@ -786,20 +786,20 @@ find_package(CURL)
 if (FETCH_STADEN)
   set(LIBSTADEN_FOUND FALSE)
 else ()
-  find_package(libstadenio)
+  find_package(libstadenio 1.14.15)
 endif()
 
 if (NOT LIBSTADENIO_FOUND)
   message("Build system will compile Staden IOLib")
   message("==================================================================")
   externalproject_add(libstadenio
     DOWNLOAD_DIR ${CMAKE_CURRENT_SOURCE_DIR}/external
-    DOWNLOAD_COMMAND curl -k -L https://github.com/COMBINE-lab/staden-io_lib/archive/v1.14.8.1.tar.gz -o staden-io_lib-v1.14.8.tar.gz &&
-      ${SHASUM} f6f30eefa478cfb708f3109a35fb6ffa0e24951d9d971985df2cef5919dd0bc3 staden-io_lib-v1.14.8.tar.gz &&
-      mkdir -p staden-io_lib-1.14.8 &&
-      tar -xzf staden-io_lib-v1.14.8.tar.gz --strip-components=1 -C staden-io_lib-1.14.8 &&
+    DOWNLOAD_COMMAND curl -k -L https://github.com/jkbonfield/io_lib/releases/download/io_lib-1-14-15/io_lib-1.14.15.tar.gz -o staden-io_lib-v1.14.15.tar.gz &&
+      ${SHASUM} 20814c4365e1e2fe6630fb11d0df370dec4c5688af3871de7f1cb0129671401e staden-io_lib-v1.14.15.tar.gz &&
+      mkdir -p staden-io_lib-1.14.15 &&
+      tar -xzf staden-io_lib-v1.14.15.tar.gz --strip-components=1 -C staden-io_lib-1.14.15 &&
       rm -fr staden-io_lib &&
-      mv -f staden-io_lib-1.14.8 staden-io_lib
+      mv -f staden-io_lib-1.14.15 staden-io_lib
     SOURCE_DIR ${CMAKE_CURRENT_SOURCE_DIR}/external/staden-io_lib
     INSTALL_DIR ${CMAKE_CURRENT_SOURCE_DIR}/external/install
     CONFIGURE_COMMAND ./configure --enable-shared=no --without-libcurl --prefix=<INSTALL_DIR> LDFLAGS=${LIBSTADEN_LDFLAGS} CFLAGS=${LIBSTADEN_CFLAGS} CC=${CMAKE_C_COMPILER} CXX=${CMAKE_CXX_COMPILER}
@@ -814,7 +814,7 @@ if (NOT LIBSTADENIO_FOUND)
   endif()
 
   set(FETCHED_STADEN TRUE)
-  set(STADEN_LIBRARIES "${GAT_SOURCE_DIR}/external/install/lib/libstaden-read.a")
+  set(STADEN_LIBRARIES "${GAT_SOURCE_DIR}/external/install/lib/libstaden-read.a;${GAT_SOURCE_DIR}/external/install/lib/libhtscodecs.a")
 endif()
 
 if (ASAN_BUILD)
@@ -906,11 +906,20 @@ set(CPACK_SOURCE_IGNORE_FILES
 
 message("CPACK_SOURCE_IGNORE_FILES = ${CPACK_SOURCE_IGNORE_FILES}")
 
+# we will use this property later
+define_property(TARGET PROPERTY COMPACT_VECTOR_DIR INHERITED 
+   BRIEF_DOCS "the path to the directory containing the compact_vector include tree"
+   FULL_DOCS "the path to the directory containing the compact_vector include tree")
+
 # Recurse into pufferfish source directory 
 # and build the library 
 set(BUILD_PUFF_FOR_SALMON TRUE)
 add_subdirectory(external/pufferfish)
 
+# make sure we know the path to compact_vector
+get_property(COMPACT_VECTOR_INCLUDE_PATH TARGET graphdump PROPERTY COMPACT_VECTOR_DIR)
+message("fetched path for compact_vector as [${COMPACT_VECTOR_INCLUDE_PATH}]")
+
 # and then the main salmon source directory
 add_subdirectory(src)
 

cmake/Modules/Findlibstadenio.cmake

@@ -10,7 +10,10 @@ find_path(STADEN_INCLUDE_DIR io_lib
 	HINTS ${STADEN_ROOT} ENV STADEN_ROOT
   PATH_SUFFIXES include)
 
-find_library(STADEN_LIBRARY NAMES staden-read libstaden-read
+find_library(STADEN_LIBRARY NAMES staden-read libstaden-read 
+  HINTS ${STADEN_ROOT} ENV STADEN_ROOT PATH_SUFFIXES lib lib64)
+
+find_library(HTSCODEC_LIBRARY NAMES htscodecs libhtscodecs
   HINTS ${STADEN_ROOT} ENV STADEN_ROOT PATH_SUFFIXES lib lib64)
 
 if(STADEN_INCLUDE_DIR)
@@ -30,7 +33,7 @@ find_package_handle_standard_args(libstadenio DEFAULT_MSG
 
 if (LIBSTADENIO_FOUND)
   message(STATUS "Staden IOLib found (include: ${STADEN_INCLUDE_DIR})")
-  set(STADEN_LIBRARIES ${STADEN_LIBRARY})
+  set(STADEN_LIBRARIES "${STADEN_LIBRARY};${HTSCODEC_LIBRARY}")
 endif()
 
 mark_as_advanced(STADEN_INCLUDE_DIR STADEN_LIBRARY)

cmake/TestSalmonQuasi.cmake

@@ -14,7 +14,10 @@ execute_process(COMMAND ${SALMON_QUASI_INDEX_CMD}
                 )
 
 if (SALMON_QUASI_INDEX_RESULT)
-    message(FATAL_ERROR "Error running ${SALMON_QUASI_INDEX_COMMAND}")
+    message(FATAL_ERROR
+            "While running: ${SALMON_QUASI_INDEX_CMD}\n"
+            "error: ${SALMON_QUASI_INDEX_RESULT}"
+            )
 endif()
 
 set(SALMON_QUANT_COMMAND ${CMAKE_BINARY_DIR}/salmon quant -i sample_salmon_quasi_index -l IU -1 reads_1.fastq -2 reads_2.fastq -o sample_salmon_quasi_quant)
@@ -23,7 +26,10 @@ execute_process(COMMAND ${SALMON_QUANT_COMMAND}
                 RESULT_VARIABLE SALMON_QUASI_QUANT_RESULT
                 )
 if (SALMON_QUASI_QUANT_RESULT)
-    message(FATAL_ERROR "Error running ${SALMON_QUASI_QUANT_RESULT}")
+    message(FATAL_ERROR
+            "While running: ${SALMON_QUANT_COMMAND}\n"
+            "error: ${SALMON_QUASI_QUANT_RESULT}"
+            )
 endif()
 
 if (EXISTS ${TOPLEVEL_DIR}/sample_data/sample_salmon_quasi_quant/quant.sf)

current_version.txt

@@ -1,3 +1,3 @@
 VERSION_MAJOR 1
-VERSION_MINOR 9
+VERSION_MINOR 10
 VERSION_PATCH 0

doc/source/conf.py

@@ -55,9 +55,9 @@
 # built documents.
 #
 # The short X.Y version.
-version = '1.9'
+version = '1.10'
 # The full version, including alpha/beta/rc tags.
-release = '1.9.0'
+release = '1.10.0'
 
 # The language for content autogenerated by Sphinx. Refer to documentation
 # for a list of supported languages.

docker/Dockerfile

@@ -6,7 +6,7 @@ MAINTAINER [email protected]
 
 ENV PACKAGES git gcc make g++ libboost-all-dev liblzma-dev libbz2-dev \
     ca-certificates zlib1g-dev libcurl4-openssl-dev curl unzip autoconf apt-transport-https ca-certificates gnupg software-properties-common wget
-ENV SALMON_VERSION 1.9.0
+ENV SALMON_VERSION 1.10.0
 
 # salmon binary will be installed in /home/salmon/bin/salmon
 

docker/build_test.sh

@@ -1,3 +1,3 @@
 #! /bin/bash
-SALMON_VERSION=1.9.0
+SALMON_VERSION=1.10.0
 docker build --no-cache -t combinelab/salmon:${SALMON_VERSION} -t combinelab/salmon:latest .

include/AlevinOpts.hpp

@@ -13,6 +13,10 @@
 
 enum class BarcodeEnd { FIVE = 5, THREE = 3 };
 enum class Sequence { BARCODE, UMI };
+// refers to which reads to use for mapping; use both for 5' libraries
+enum class ReadsToMap { USE_BOTH, USE_FIRST, USE_SECOND };
+// refers to which reads map to transcripts and to which transcripts they map to
+enum class PairingStatus { UNPAIRED_LEFT, UNPAIRED_RIGHT, PAIRED_FR, PAIRED_RF };
 
 template <class protocolT>
 struct AlevinOpts {
@@ -82,11 +86,11 @@ struct AlevinOpts {
   // initialize EM with uniform prior
   bool initUniform;
   //number of cells
-  uint32_t numCells;
+  uint64_t numCells;
   // minimum number of CB to use for low confidence region
-  uint32_t lowRegionMinNumBarcodes;
+  uint64_t lowRegionMinNumBarcodes;
   // maximum number of barcodes to use
-  uint32_t maxNumBarcodes;
+  uint64_t maxNumBarcodes;
   // number of bootstraps to perform
   uint32_t numBootstraps;
   // number of gibbs samples to perform
@@ -115,27 +119,27 @@ struct AlevinOpts {
   boost::filesystem::path bfhFile;
 
   //meta-info related tags
-  uint32_t totalReads;
-  uint32_t totalUsedReads;
-  uint32_t readsThrown;
+  uint64_t totalReads;
+  uint64_t totalUsedReads;
+  uint64_t readsThrown;
 
-  uint32_t totalCBs;
-  uint32_t totalUsedCBs;
+  uint64_t totalCBs;
+  uint64_t totalUsedCBs;
 
-  uint32_t kneeCutoff;
-  uint32_t intelligentCutoff;
-  uint32_t totalLowConfidenceCBs;
-  uint32_t numFeatures;
-  uint32_t numNoMapCB;
+  uint64_t kneeCutoff;
+  uint64_t intelligentCutoff;
+  uint64_t totalLowConfidenceCBs;
+  uint64_t numFeatures;
+  uint64_t numNoMapCB;
 
-  uint32_t eqReads;
-  uint32_t noisyUmis;
+  uint64_t eqReads;
+  uint64_t noisyUmis;
   double mappingRate;
   double keepCBFraction;
   double vbemNorm;
 
-  uint32_t totalDedupUMIs;
-  uint32_t totalExpGenes;
+  uint64_t totalDedupUMIs;
+  uint64_t totalExpGenes;
 };
 
 #endif // ALEVIN_OPTS_HPP

include/SalmonConfig.hpp

@@ -26,9 +26,9 @@
 
 namespace salmon {
 constexpr char majorVersion[] = "1";
-constexpr char minorVersion[] = "9";
+constexpr char minorVersion[] = "10";
 constexpr char patchVersion[] = "0";
-constexpr char version[] = "1.9.0";
+constexpr char version[] = "1.10.0";
 constexpr uint32_t indexVersion = 5;
 constexpr char requiredQuasiIndexVersion[] = "p7";
 } // namespace salmon