alexanderdilthey / kmerificator

kMerificator

kmerificator's Issues

Future feature enhancement: update README, add --help flag to sourcecode

I should probably add a --help flag to this code, so users could see the parameters and parameter meanings.

This should be detailed in the README as well, kMerificator.cpp:

if(arguments.at(i) == "--vcf")
{
	vcfFile = arguments.at(i+1);
}

if(arguments.at(i) == "--referenceGenome")
{
	referenceGenome = arguments.at(i+1);
}

if(arguments.at(i) == "--deBruijnGraph")
{
	deBruijnGraph = arguments.at(i+1);
}

if(arguments.at(i) == "--outputDirectory")
{
	outputDirectory = arguments.at(i+1);
}


if(arguments.at(i) == "--cortex_height")
{
	cortex_height = Utilities::StrtoI(arguments.at(i+1));
}

if(arguments.at(i) == "--cortex_width")
{
	cortex_width = Utilities::StrtoI(arguments.at(i+1));
}

if(arguments.at(i) == "--k")
{
	kMer_size = Utilities::StrtoI(arguments.at(i+1));
}

if(arguments.at(i) == "--threads")
{
	threads = Utilities::StrtoI(arguments.at(i+1));
}

if(arguments.at(i) == "--onlyPASS")
{
	onlyPASS = Utilities::StrtoI(arguments.at(i+1));
}

if(arguments.at(i) == "--maximumHaplotypePairs")
{
	maximumHaplotypePairs = Utilities::StrtoI(arguments.at(i+1));
}

if(arguments.at(i) == "--regions")
{
	regions = arguments.at(i+1);
}	

Installing on Mac OS and Ubuntu, future README revision

This is more of a note to my self for the future, but I suppose one could revise the README in the future.

In order to compile on a Mac (circa 2020, v10.15.2):

Recall gcc on Mac isn't gcc. And the llvm libraries installed via brew do not contain standard C++ libraries.
https://stackoverflow.com/questions/14128298/using-homebrew-gcc-and-llvm-with-c-11

So the Makefile I wrote up uses the following:

## boost libraries installed via brew
INC_BOOST = /usr/local/Cellar/boost/1.72.0/include
LIB_BOOST = /usr/local/Cellar/boost/1.72.0/lib
INCS = -I$(INC_BOOST) 
LIBS = -L$(LIB_BOOST) -lboost_system -lboost_filesystem

### explicitly set these
DIR_OBJ = /Users/evanbiederstedt/kMerificator/obj
DIR_BIN = /Users/evanbiederstedt/kMerificator/bin

## I suspect I've named these incorrectly...
CPP = clang++ -std=c++11 -stdlib=libc++
CPPFLAGS = -I/usr/local/opt/llvm/include -fopenmp
LDFLAGS = -L/usr/local/opt/llvm/lib
COPTS  = -ggdb -O2 -fopenmp -std=gnu++0x -fstack-protector-all
## CFLAGS = 
COMPILE = $(CPP) $(COPTS) $(INCS) $(CPPFLAGS) $(LDFLAGS)

---

Here is the entire Makefile, which worked:

# LIBRARY SETTINGS - SET AS NECESSARY


INC_BOOST = /usr/local/Cellar/boost/1.72.0/include
LIB_BOOST = /usr/local/Cellar/boost/1.72.0/lib
INCS = -I$(INC_BOOST) 
LIBS = -L$(LIB_BOOST) -lboost_system -lboost_filesystem


MKDIR_P = mkdir -p

.PHONY: directories
	
# END LIBRARY SETTINGS

#
# object and binary dirs
#

DIR_OBJ = /Users/evanbiederstedt/kMerificator/obj
DIR_BIN = /Users/evanbiederstedt/kMerificator/bin

CPP = clang++ -std=c++11 -stdlib=libc++
CPPFLAGS = -I/usr/local/opt/llvm/include -fopenmp
LDFLAGS = -L/usr/local/opt/llvm/lib
CXX    = g++
COPTS  = -ggdb -O2 -fopenmp -std=gnu++0x -fstack-protector-all
## CFLAGS = 
COMPILE = $(CPP) $(COPTS) $(INCS) $(CPPFLAGS) $(LDFLAGS)

VPATH = Data:Graph:NextGen:hash:hash/deBruijn:hash/sequence:GraphAligner:GraphAlignerUnique:readFilter
        
OBJS = \
        $(DIR_OBJ)/Utilities.o \
        $(DIR_OBJ)/DeBruijnGraph.o \
        $(DIR_OBJ)/DeBruijnElement.o \
        $(DIR_OBJ)/basic.o \
        $(DIR_OBJ)/binarykMer.o \
        $(DIR_OBJ)/Hsh.o \
        $(DIR_OBJ)/Validator.o \
#
# list executable file names
#
EXECS = kMerificator

OUT_DIR = ../obj ../bin

directories: ${OUT_DIR}


#
# compile and link
#
default:
	@echo
	@echo " to build:"
	@echo "    make all"
	@echo
	@echo " to clean:"
	@echo "    make clean"
	@echo "    make realclean"
	@echo

all: directories $(EXECS)

$(EXECS): $(OBJS)
	$(foreach EX, $(EXECS), $(COMPILE) $(EX).cpp -c -o $(DIR_OBJ)/$(EX).o;)
	$(foreach EX, $(EXECS), $(COMPILE) $(OBJS) $(DIR_OBJ)/$(EX).o -o $(DIR_BIN)/$(EX) $(LIBS);)

$(DIR_OBJ)/%.o: %.cpp %.h
	$(COMPILE) $< -c -o $@


#
# odds and ends
#
clean:
	/bin/rm $(DIR_OBJ)/*

realclean: clean
	/bin/rm $(DIR_BIN)/*

${OUT_DIR}:
	${MKDIR_P} ${OUT_DIR}

README additions, outputs

Additions to README

Method: With / without VCF

• Total characters: Length of the optimal pair of considered haplotypes (the thing assumes diploidy, hence here 2 x chromosome length)

• Total non-gap characters: Non-gap characters in the optimal pair of considered haplotypes

• kMers: Number of kmers in the optimal pair of haplotypes

• kMers invalid: Of these, invalid kmers (e.g. weird characters)

• kMers present: Of the valid kmers, kMers recovered from the de Bruijn graph
  Unweighted optimality: Proportion of kmers in the optimal pair of considered haplotypes classified as 'recovered'

• Coverage-weighted optimality: ??

---

- Level: Row ID, 0-indexed
- ReferenceCoordinate: location of basepair, 1-index
- kMersInduced:
- kMersInvalid:
- kMersOK:
- InducedkMersInReference:
- DiploidChromotype:
- ChromotypeLostPhase:
- GapsAtLevel: