Platanus is a novel de novo sequence assembler that can reconstruct genomic sequences of
highly heterozygous diploids from massively parallel shotgun sequencing data.
The latest version is 1.2.4.
To cite Platanus, please use the following:
Kajitani R, Toshimoto K, Noguchi H, Toyoda A, Ogura Y, Okuno M, Yabana M, Harada M, Nagayasu E, Maruyama H, Kohara Y, Fujiyama A, Hayashi T, Itoh T, “Efficient de novo assembly of highly heterozygous genomes from whole-genome shotgun short reads”. Genome Res. 2014 Aug;24(8):1384-95. doi: 10.1101/gr.170720.113. [abstract | full text]
Comments
Platanus README ***** VERSION ***** 1.2.3 ***** DESCRIPTION ***** Platanus is a de novo assembler designed to assemble high-throughput data. It can handle highly heterozygotic samples. The following is the assembly outline. First, it constructs contigs using the algorithm based on de Bruijn graph. Second, the order of contigs is determined according to paired-end (mate-pair) data and constructs scaffolds. Finally, paired-end reads localized on gaps in scaffolds are assembled and gaps are closed. ***** REQUIREMENTS ***** GCC (version >= 4.4) OpenMP ***** INSTALLATION ***** Command: > tar xzfv Platanus_<version>.tar.gz > cd Platanus_<version> > make > cp platanus <installation_path> The command of c++ compiler can be specified by editing "CXX" in Makefile. e.g. #CXX = g++ CXX = g++44 ***** SYNOPSIS ***** > platanus assemble -f PE.fa 2>ass_log.txt > platanus scaffold -c out_contig.fa -b out_contigBubble.fa -ip1 PE.fa 2>sca_log.txt > platanus gap_close -c out_scaffold.fa -ip1 PE.fa 2>gap_log.txt ***** USAGE ***** COMMON OPTIONS -------------------------------------------------------------------------------- -t INT : Number of threads (<= 100, default 1) -o STR : Prefix of output files (default out, length <= 200) -tmp DIR : directory for temporary files (default . (working directory)) -------------------------------------------------------------------------------- platanus assemble [OPTIONS] 2>log -------------------------------------------------------------------------------- Constructs contigs using the algorithm based on de Bruign graph. INPUT OPTIONS -f FILE1 [FILE2 ...]: Reads file (fasta or fastq format, number <= 100) The file format is automatically detected. Quality values are not used. OTHER OPTIONS -k INT : Initial k-mer size (default 32) For low-coverage data, small INT is recommended. -s INT : Step size of k-mer extension (>= 1, default 10) Smaller INT increase time and may enhance accuracy of contigs. -n INT : Initial k-mer coverage cutoff (default 0, 0 means auto) When "-n 0", the value depends on k-mer occurrence distribution. If k-mer occurrence distribution is abnormal (Ex. contaminated, transcriptome, metagenome, and so on), the value should be set manually. -c INT : Minimum k-mer coverage (default 2) Through assembly, k-mer size increases and coverage cutoff decreases. The coverage cutoff does not fall below INT. -a FLOAT: K-mer extension safety level (default 10.0) Smaller FLOAT increases the final k-mer size. If you want to extend contigs at the cost of accuracy, set small FLOAT (Ex. -a 5.0). -u FLOAT: Maximum difference for bubble crush (identity, default 0.1) Larger FLOAT increases the number of bubbles merged. If heterozygosity of the sample is high, large FLOAT may be suitable (Ex. -u 0.2). -d FLOAT: Maximum difference for branch cutting (coverage ratio, default 0.5) Smaller FLOAT increase the accuracy. If error rate is low, small FLOAT may be suitable (Ex. -d 0.3). -m INT : Memory limit (GB, >= 1, default 16) Programs attempt to make memory usage smaller than INT(GB). If memory usage exceed the limit, programs warn but continue. OUTPUT FILES PREFIX_contig.fa : assembled contiguous sequences PREFIX_contigBubble.fa: merged and removed bubble sequences PREFIX_kmerFrq.tsv : occurrence distribution of k-mers(k is specified by -k option) -------------------------------------------------------------------------------- platanus scaffold [OPTIONS] 2>log -------------------------------------------------------------------------------- Map paired-end (mate-pair) reads on contigs, determine the order of contigs and construct scaffolds. INPUT OPTIONS -c FILE1 [FILE2 ...] : Contig_file (fasta format) String "cov" in title lines are detected and following numbers are used as coverage. Even if a title does not consits of string "cov", Platanus can process the file. Ex. out_contig.fa -b FILE1 [FILE2 ...] : Bubble_seq_file (fasta format) Ex. out_contigBubble.fa -ip{INT} PAIR1 [PAIR2 ...] : Lib_id inward_pair_file (reads in 1 file, fasta or fastq) Ex. -ip1 lib1.fa -IP{INT} FWD1 REV1 [FWD2 REV2 ...] : Lib_id inward_pair_files (reads in 2 files, fasta or fastq) Ex. -IP1 lib1_1.fa lib1_2.fa -op{INT} PAIR1 [PAIR2 ...] : Lib_id outward_pair_file (reads in 1 file, fasta or fastq) Ex. -op1 lib1.fa -OP{INT} FWD1 REV1 [FWD2 REV2 ...] : lib_id outward_pair_files (reads in 2 files, fasta or fastq) Ex. -OP1 lib1_1.fa lib1_2.fa The file format is automatically detected. see "***** NOTE ***** Paired-end (Mate-pair) input" below OTHER OPTIONS -n{INT} INT: Lib_id Minimum_insert_size Platanus automatically estimates insert size of each library. If a library consists of many short-insert pairs (junks), the insert size will be underestimated. Pairs in the library (Lib_id=INT1) that infer short insert size (< INT2) are discarded and estimated insert size must be > INT2. -a{INT} INT : lib_id average_insert_size Fixed average insert size (INT) is used instead of auto estimation. -d{INT} INT : lib_id SD_insert_size Fixed SD of insert size (INT) is used instead of auto estimation. -s INT : Mapping seed length (default 32) Seed length must be larger than reads length. Smaller INT decrease speed. -v INT : Minimum overlap length (default 32) If adjacent contigs have overlap (length >= INT) and properly close to each other, the contigs are joined. -l INT : Minimum number of link (default 3) Platanus first estimates the threshold of link (number) and makes scaffolds, then decreases the threshold to INT and extends scaffolds. -u FLOAT : Maximum difference for bubble crush (identity, default 0.1) Larger FLOAT increases the number of bubbles merged. If heterozygosity of the sample is high, large FLOAT may be suitable (Ex. -u 0.2). OUTPUT FILES PREFIX_scaffold.fa : assembled sequences that include gaps('N's mean gaps) PREFIX_scaffoldBubble.fa : removed bubble sequences PREFIX_scaffoldComponent.tsv: the information about composition of scaffolds (i.e. which contigs constitute a scaffold) -------------------------------------------------------------------------------- platanus gap_close [OPTIONS] 2>log -------------------------------------------------------------------------------- Map paired-end(mate-pair) reads on scaffolds, assemble reads localized on gaps and close gaps. INPUT OPTIONS -c FILE1 [FILE2 ...] : Scaffold_file (fasta format) Ex. out_scaffold.fa -ip{INT} PAIR1 [PAIR2 ...] : Lib_id inward_pair_file (reads in 1 file, fasta or fastq) Ex. -ip1 lib1.fa -IP{INT} FWD1 REV1 [FWD2 REV2 ...] : Lib_id inward_pair_files (reads in 2 files, fasta or fastq) Ex. -IP1 lib1_1.fa lib1_2.fa -op{INT} PAIR1 [PAIR2 ...] : Lib_id outward_pair_file (reads in 1 file, fasta or fastq) Ex. -op1 lib1.fa -OP{INT} FWD1 REV1 [FWD2 REV2 ...] : lib_id outward_pair_files (reads in 2 files, fasta or fastq) Ex. -OP1 lib1_1.fa lib1_2.fa The file format is automatically detected. see "***** NOTE ***** Paired-end (Mate-pair) input" below OTHER OPTIONS -s INT : Mapping seed length (default 32) Seed length must be larger than reads length. Smaller INT decrease speed. -v INT : Minimum overlap length (default 32) Smaller INT increase the number of gaps closed (Ex. -v 20). -e FLOAT: Maximum error rate of overlap (identity, default 0.05) Larger FLOAT increase the number of gaps closed (Ex. -e 0.1). OUTPUT FILE PREFIX_gapClosed.fa: gap-closed scaffold sequences -------------------------------------------------------------------------------- ****** NOTE ****** Paired-end (Mate-pair) input --------------------------------------------------------------------------------- "platanus scaffold" and "platanus gap_close" require Paired-end (Mate-pair) libraries. Paired-end libraries are classified into "Inward-pair" and "Outward-pair" according to the sequence direction. Libraries that have the same insert size are given the same Lib_id (INT). Inward-pair data (often called "Paired-end", accepted in options "-ip" or "-IP"): FWD ---> 5' -------------------- 3' 3' -------------------- 5' <--- REV Outward-pair data (often called "Mate-pair", accepted in options "-op" or "-OP"): ---> REV 5' -------------------- 3' 3' -------------------- 5' FWD <--- EXAMPLE INPUT Inward-pair(Insert = 300bp, reads in 1 file): PE300_1_pair.fa PE300_2_pair.fa Inward-pair(Insert = 500bp, reads in 1 file): PE500_pair.fa Outward-pair(Insert = 2kbp, reads in 2 files) : MP2k_fwd.fa MP2k_rev.fa OPTIONS -ip1 PE300_1_pair.fa PE300-2_pair.fa \ -ip2 PE500_pair.fa \ -OP3 MP2k_fwd.fa MP2k_rev.fa -------------------------------------------------------------------------------- ***** AUTHOR ***** Rei Kajitani at the Tokyo Institute of Technology wrote key source codes. <platanus@bio.titech.ac.jp>