config.md

• project: project name. The pipeline makes a folder named project to contain all output files.
• reads: the read file. It should be a .fasta or .fastq file. It can also be a .txt file which contains the full paths of all read files.
• genome_size: estimation of genome size.
• threads: CPU threads for each job.
• memory: memory for each job.
• cleanup: 1 delete temporary files, 0 do not delete temporary files
• grid: cluster system. It supports PBS, SGE, LSF and Slurm systems.
  • auto: automatically detecting the type of cluster system.
  • pbs: PBS systems.
  • sge: SGE systems.
  • lsf: LSF systems.
  • slurm:Slurm systems.
  • local: do not use cluster system.
• grid_options: It is used to add additional options for submitting jobs.

Preparing

The pipeline extracts genome_size*prep_output_coverage longest reads from reads for the next steps.

• prep_min_length: minimum length of reads
• prep_output_coverage: coverage of extracted reads from raw reads.

Correcting

Long nosiy reads are corrected in this step.

• corr_iterate_number: number of rounds for correction.

• corr_block_size: the reads are split into multiple blocks for parallel processing.

• corr_rd2rd_options: parameters for minimap2 to find the candiate overlaps between raw reads.

• corr_filter_options: parameters for filtering low-quality overlaps.

  • filter0: filter parameters.
    • l: minimun read length
    • al: minimum overlap length
    • alr: minimum ratio of overlap length to read length.
    • oh: maximum overhang length.
    • ohr: maximum ratio of overhang length to read length.
    • aal: the overlap is not filtered if the length exceeds this value.
    • aalr: the overlap is not filtered if ratio of overlap length to read length exceeds this value.

• corr_correct_options: parameters for correcting raw reads.

  • aligner: local alignment algorithm. It can be set to diff and edlib.
  • min_coverage: minimum coverage of base for correcting.
  • score: scoring method for selecting supporting reads. weight for diploid datasets. count for haploid datasets. lc: minimum coverage for detecting haplotype branch in the POA graph.
  • candidate: parameters for selecting candidate overlaps for local alignment. n: maxinum number for implementing local alignment. f: maximum number of consecutive local alignment failures.
  • min_identity: minimum identity of alignment between the tempalte read and the supporting read.
  • min_local_identity: minimum local identity where the local window length is 1000.
  • filter0: parameters for filtering overlaps before implementing local alignment. See corr_filter_options.
  • filter1: parameters for filtering overlaps after implementing local alignment. See corr_filter_options.

• corr_output_coverage: coverage for extracting reads from corrected reads. The pipeline extracts genome_size*corr_output_coverage longest reads for the next steps.

Algining

The pipeline finds the the overlaps between corrected reads.

• align_block_size: the reads are split into multiple blocks for parallel processing.
• align_rd2rd_options: parameters for minimap2 to find the candiate overlaps between corrected reads.
• align_filter_options: parameters for filtering low-quality overlaps and extend the overlaps to the ends of the reads.
  • filter0: filter parameter beforing local alignments. (see corr_filter_options)
  • task: extend for performing local alignment on overhangs.
  • filter1: filter parameters aftering local alignment. (see corr_filter_options)

The first round of assembly

The pipeline generated haplotype-collapsed contigs.

• asm1_assemble_options: parameters for the first assembly.
  • max_trivial_length: pipeline uses the value to determine which edges in the complex region of the graph are trivial.
  • min_identity: minimum identity of the edges in the assembly graph.
  • min_coverage: minimum coverage of the reads.
  • min_contig_length: mininum contig length for output.
  • contig_format: prialt output primary/alternate format contigs, dual output dual format contigs, prialt,dual output both format contigs.
  • reducer0: parameters for graph reduction best: best overlap graph algorithm. cmp: parameters for comparing two edges. phase: using SNP alleles to phase adjacent bubble structures. sc: minimum support for a SNP allele.
  • reducer1: parameters for graph reduction
    • spur: removing spur paths. length maximum spur length, nodesize: maximum node size in graph of spur

Phasing

The pipeline identifies the inconsistent overlaps. The reads are mapped to the haplotype-collapsed contigs. PECAT use built-in method or clair3 to identity the SNPs. According to SNP allels in reads, PECAT identifies and remove inconsistent overlaps.

• phase_method: using built-in method or clair3. 0 built-in method, 1 clair3, 2 both methods.

• phase_use_reads: using raw reads(0) or corrected reads(1) for phasing.

• phase_rd2ctg_options: parameters for minimap2 to find the alignments bewteen the haplotype-collapsed contigs and the reads.

• phase_phase_options: parameters for detecting inconsistent overlaps.

  • coverage: parameters for coverage for detecting SNP sites.
    • lc: minimum coverage.
  • phase_options parameters for dececting inconsistent overlaps.
    • icr: minimum ratio of different SNP alleles to all SNP alleles for identifing the pair of inconsistent reads.
    • icc: minimum different SNP alleles for identifing the pair of inconsistent reads.
    • sc: minimum support for a SNP allele.

• phase_filter_options: parameters for filtering out inconsistent overlaps.

  • threshold: maxinum offset between overlaps detected by minimap2 and inconsistent overlaps detected in the phasing step. 1000 for corrected reads and -1 for raw reads.
  • rate: maximum rate of offset to read length.

• phase_clair3_command: if using singularity, it can be set singularity exec --containall -B ``pwd -P``:``pwd -P`` clair3_v0.1-r12.sif /opt/bin/run_clair3.sh

• phase_clair3_options: parameters passed to clair3.

• phase_clair3_rd2ctg_options: similar to phase_rd2ctg_options.

• phase_clair3_phase_options: similar to phase_phase_options.

• phase_clair3_use_reads: similar to phase_use_reads.

• phase_clair3_filter_options: similar to phase_filter_options.

The second round of assembly

After removing the inconsistent overlaps, the pipeline reassembles the corrected reads again.

• asm2_assemble_options: parameters for the second round of assembly. See asm1_assemble_options.

Polishing

By default, we only use racon to polish contigs. If polish_medaka=1 is set, we further polish the output of racon with medaka.

Parameters for using racon.

• polish_use_reads: using raw reads(0) or corrected reads(1) for polishing.
• polish_map_options: parameters for minimap2 to find the alignments bewteen the second assembly and the reads.
• polish_filter_options: parameters for filtering low-quality alignments (see corr_filter_options).
• polish_cns_options: parameters for racon to polish contigs.

Parameters for using medaka.

• polish_medaka: whether to use medaka to polish contigs.
• polish_medaka_command: medaka command. If using singularity, it can be set singularity exec --containall -B `pwd -P`:`pwd -P` medaka_1.7.2--aa54076.sif medaka
• polish_medaka_map_options: parameters for minimap2 to find the alignments bewteen the outputs of racon and the reads.
• polish_medaka_cns_options: parameters for medaka.