hifiasm/hifiasm.1

.TH hifiasm 1 "22 August 2021" "hifiasm-0.16.0 (r369)" "Bioinformatics tools"

.SH NAME
.PP
hifiasm - haplotype-resolved de novo assembler for PacBio Hifi reads.

.SH SYNOPSIS

* Assemble HiFi reads:
.RS 4
.B hifiasm
.RB [ -o
.IR prefix ]
.RB [ -t
.IR nThreads ]
.RB [ -z
.IR endTrimLen ]
.R [options]
.I input1.fq
.RI [ input2.fq
.R [...]]
.RE

* Trio binning assembly with yak dumps:
.RS 4
.B yak count
.B -o
.I paternal.yak
.B -b37
.RB [ -t
.IR nThreads ]
.RB [ -k
.IR kmerLen ]
.I paternal.fq.gz
.br
.B yak count
.B -o
.I maternal.yak
.B -b37
.RB [ -t
.IR nThreads ]
.RB [ -k
.IR kmerLen ]
.I maternal.fq.gz
.br
.B hifiasm
.RB [ -o
.IR prefix ]
.RB [ -t
.IR nThreads ]
.R [options]
.B -1
.I paternal.yak
.B -2
.I maternal.yak
.I child.hifi.fq.gz
.RE

.SH DESCRIPTION
.PP
Hifiasm is an ultrafast haplotype-resolved de novo assembler for PacBio
Hifi reads. Unlike most existing assemblers, hifiasm starts from uncollapsed
genome. Thus, it is able to keep the haplotype information as much as possible.
The input of hifiasm is the PacBio Hifi reads in fasta/fastq format, and its
outputs consist of multiple types of assembly graphs in GFA format.


.SH OPTIONS

.SS General options

.TP 10
.BI -o \ FILE
Prefix of output files [hifiasm.asm]. For detailed description of all assembly
graphs, please see the
.B OUTPUTS
section of this man-page. 

.TP 10
.BI -t \ INT
Number of CPU threads used by hifiasm [1]. 

.TP
.BI -h
Show help information.

.TP
.BI --version
Show version number. 


.SS Error correction options

.TP 10
.BI -k \ INT
K-mer length [51]. This option must be less than 64.

.TP
.BI -w \ INT
Minimizer window size [51].

.TP
.BI -f \ INT
Number of bits for bloom filter; 0 to disable [37]. This bloom filter is used
to filter out singleton k-mers when counting all k-mers. It takes
.RI 2^( INT -3)
bytes of memory. A proper setting saves memory. 
.BR -f37 
is recommended for human
assembly. For small genomes, use 
.BR -f0 
to disable the initial bloom filter 
which takes 16GB memory at the beginning. For genomes much larger 
than human, applying 
.BR -f38 
or even 
.BR -f39 
is preferred to save memory on k-mer counting.

.TP
.BI -D \ INT
Drop k-mers occurring 
.I >FLOAT*coverage 
times [5.0]. 
Hifiasm discards these high-frequency k-mers 
during error correction to reduce running time. 
The 
.I coverage 
is determined automatically 
by hifiasm based on k-mer plot, representing 
homozygous read coverage. Raising this option 
may improve the resolution of repetitive regions 
but takes longer time.

.TP
.BI -N \ INT
Consider up to 
.I max(-D*coverage,-N) 
overlaps for each oriented read [100]. 
The 
.I coverage 
is determined automatically 
by hifiasm based on k-mer plot, representing 
homozygous read coverage. Raising this option may 
improve the resolution of repetitive regions but 
takes longer time.


.TP
.BI -r \ INT
Rounds of haplotype-aware error corrections [3]. 
This option affects all outputs of hifiasm. 
Odd rounds of correction are preferred in practice.

.TP
.BI -z \ INT
Length of adapters that should be removed [0]. This option remove
.I INT
bases from both ends of each read.
Some old Hifi reads may consist of
short adapters (e.g., 20bp adapter at one end). For such data, trimming short adapters would 
significantly improve the assembly quality.

.TP
.BI --max-kocc \ INT
Employ k-mers occurring < 
.IR INT 
times to rescue repetitive overlaps [2000]. 
This option may improve the resolution of repeats.

.TP
.BI --hg-size \ INT (k/m/g)
Estimated haploid genome size used for inferring read coverage [auto]. 
This option is used to get accurate homozygous read coverage during 
error correction. Common suffices are required, for example, 100m or 3g.

.TP
.BI --min-hist-cnt \ INT
When analyzing the k-mer spectrum, ignore counts below
.IR INT 
[5]. For very low coverage of HiFi data, set smaller 
value for this option.

.SS Assembly options

.TP
.BI -a \ INT
Rounds of assembly graph cleaning [4]. This option is used with
.B -x
and
.BR -y .
Note that unlike
.BR -r ,
this option does not affect error corrected reads and all-to-all overlaps.


.TP
.BI -m \ INT
Maximal probing distance for bubble popping when generating primary/alternate
contig graphs [10000000]. Bubbles longer than
.I INT
bases will not be popped. For detailed description of these graphs, please see the
.B OUTPUTS
section of this man-page. 

.TP
.BI -p \ INT
Maximal probing distance for bubble popping when generating haplotype-resolved processed unitig graph
without small bubbles [0]. Bubbles longer than
.I INT
bases will not be popped. Small bubbles might be caused by somatic mutations or noise in data. 
Please note that hifiasm automatically pops small bubbles based on coverage, 
which can be tweaked by 
.BR --hom-cov .
For detailed description of this graph, please see the
.B OUTPUTS
section of this man-page. 

.TP
.BI -n \ INT
A unitig is considered small if it is composed of less than 
.I INT
reads [3]. Hifiasm may try to remove small unitigs at various steps.

.TP
.BI -x \ FLOAT1, -y \ FLOAT2
Max and min overlap drop ratio [0.8, 0.2]. This option is used with
.BR -a .
Given a node
.I N
in the assembly graph, let max(N)
be the length of the longest overlap of
.I N.
Hifiasm iteratively drops overlaps of
.I N
if their length / max(N)
are below a threshold controlled by
.B -x
and
.BR -y .
Hifiasm applies
.B -a
rounds of short overlap removal with an increasing threshold between
.I FLOAT1
and
.I FLOAT2.

.TP
.BI -i
Ignore error corrected reads and overlaps saved in
.IR prefix .*.bin
files so that hifiasm will start again from scratch.
Apart from assembly graphs, hifiasm also outputs three binary files
that save all overlap information during assembly step.
With these files, hifiasm can avoid the time-consuming all-to-all overlap calculation step,
and do the assembly directly and quickly.
This might be helpful when users want to get an optimized assembly by multiple rounds of experiments
with different parameters.

.TP
.BI -u
Disable post-join step for contigs which may improve N50. 
The post-join step of hifiasm improves contig N50 but may introduce misassemblies.

.TP
.BI --hom-cov \ INT
Homozygous read coverage inferred automatically in default. This option affects different types of outputs, 
including Hi-C phased assembly and HiFi-only assembly. 

.TP
.BI --pri-range \ INT1[,INT2]
Min and max coverage cutoff of primary contigs.
Keep contigs with coverage in this range at p_ctg.gfa. 
Inferred automatically in default.
If 
.I INT2 
is not specified, it is set to infinity. 
Set -1 to disable.

.TP
.BI --lowQ \ INT
Output contig regions with >=INT% inconsistency to the bed file 
with suffix
.B lowQ.bed
[70]. Set 0 to disable. 


.TP
.BI --b-cov \ INT
Break contigs at potential misassemblies with <INT-fold coverage [0].
Work with 
.B --m-rate.
Set 0 to disable. 

.TP
.BI --h-cov \ INT
Break contigs at potential misassemblies with >INT-fold coverage [-1].
Work with 
.B --m-rate.
Set -1 to disable. 

.TP
.BI --m-rate \ FLOAT
Break contigs with <=FLOAT*coverage exact overlaps [0.75].
Only work with 
.B --b-cov
and
.B --h-cov.

.TP
.BI --primary
Output a primary assembly and an alternate assembly.
Hifiasm outputs two balanced assemblies and a primary
assembly in default. Enable this option or 
.B -l0
outputs a primary assembly and an alternate assembly.


.SS Trio-partition options

.TP 10
.BI -1 \ FILE
K-mer dump generated by
.B yak count
from the paternal/haplotype1 reads []

.TP
.BI -2 \ FILE
K-mer dump generated by
.B yak count
from the maternal/haplotype2 reads []

.TP
.BI -3 \ FILE
List of paternal/haplotype1 read names []

.TP
.BI -4 \ FILE
List of maternal/haplotype2 read names []

.TP
.BI -c \ INT
Lower bound of the binned k-mer's frequency [2]. When doing trio binning, 
a k-mer is said to be differentiating if it occurs >=
.B -d
times in one sample 
but occurs <
.B -c
times in the other sample.

.TP
.BI -d \ INT
Upper bound of the binned k-mer's frequency [5]. When doing trio binning, 
a k-mer is said to be differentiating if it occurs >=
.B -d
times in one sample 
but occurs <
.B -c
times in the other sample.

.TP
.BI --t-occ \ INT
Forcedly remove unitig including >
.I INT 
unexpected haplotype-specific reads 
without considering graph topology [60].


.SS Purge-dups options

.TP 10
.BI -l \ INT
Level of purge-dup. 0 to disable purge-dup, 1 to only purge contained haplotigs, 
2 to purge all types of haplotigs, 3 to purge all types of haplotigs in most aggressive way. 
In default, [3] for non-trio assembly, [0] for trio assembly.
For trio assembly, only level 0 and level 1 are allowed.

.TP
.BI -s \ FLOAT
Similarity threshold for duplicate haplotigs that should be purged [0.75 for 
.BR -l1/-l2 , 
0.55 for 
.BR -l3 ].

.TP
.BI -O \ FLOAT
Min number of overlapped reads for duplicate haplotigs that should be purged [1].

.TP
.BI --purge-max \ INT
Coverage upper bound of Purge-dups, which is inferred automatically in default.
If the coverage of a contig is higher than this bound, don't apply Purge-dups. 
Larger value makes assembly more contiguous but may collapse repeats or segmental duplications.

.TP
.BI   --n-hap \ INT
Assumption of haplotype number. If it is set to >2, the quality of 
primary assembly for polyploid genomes might be improved.



.SS Hi-C-partition options [experimental, not stable]

.TP
.BI --h1 \ FILEs 
File names of input Hi-C R1 [r1_1.fq,r1_2.fq,...].

.TP
.BI --h2 \ FILEs 
File names of input Hi-C R2 [r2_1.fq,r2_2.fq,...].

.TP
.BI --n-weight \ INT 
Rounds of reweighting Hi-C links [3]. Increasing this may improves
phasing results but takes longer time.

.TP
.BI --n-perturb \ INT 
Rounds of perturbation [10000]. Increasing this may improves
phasing results but takes longer time.

.TP
.BI --f-perturb \ FLOAT 
Fraction to flip for perturbation [0.1]. Increasing this may improves
phasing results but takes longer time.

.TP
.BI --l-msjoin \ INT 
Detect misjoined unitigs of >=
.I INT 
in size; 0 to disable [500000].

.TP
.BI --seed \ INT 
RNG seed [11].

.SH OUTPUTS

.PP
In general, hifiasm generates the following assembly graphs in the GFA format:

.RS 2
.TP 2
*
.IR prefix .r_utg.gfa:
haplotype-resolved raw unitig graph. This graph keeps all haplotype information.

.TP
*
.IR prefix .p_utg.gfa:
haplotype-resolved processed unitig graph without small bubbles.  Small bubbles
might be caused by somatic mutations or noise in data, which are not the real
haplotype information. Hifiasm automatically pops such small bubbles based on coverage. 
The option 
.BR --hom-cov 
affects the result.
In addition, the option 
.BR -p 
forcedly pops bubbles.

.TP
*
.IR prefix .p_ctg.gfa:
assembly graph of primary contigs. This graph includes a complete assembly with 
long stretches of phased blocks.

.TP
*
.IR prefix .a_ctg.gfa:
assembly graph of alternate contigs. This graph consists of all contigs that
are discarded in primary contig graph.

.TP
*
.IR prefix .*hap*.p_ctg.gfa:
phased contig graph. This graph keeps the phased assembly.

.RE

.PP
Hifiasm outputs
.B *.r_utg.gfa
and
.B *.p_utg.gfa
in any cases.
Specifically, hifiasm outputs the following assembly graphs
with trio-binning options:

.RS 2
.TP 2
*
.IR prefix .dip.hap1.p_ctg.gfa:
fully phased paternal/haplotype1 contig graph keeping the phased
paternal/haplotype1 assembly.

.TP
*
.IR prefix .dip.hap2.p_ctg.gfa:
fully phased maternal/haplotype2 contig graph keeping the phased
maternal/haplotype2 assembly.
.RE

.PP
With Hi-C partition options, hifiasm outputs:

.RS 2
.TP 2
*
.IR prefix .hic.p_ctg.gfa:
assembly graph of primary contigs.

.TP
*
.IR prefix .hic.hap1.p_ctg.gfa:
fully phased contig graph where each contig is fully phased.

.TP
*
.IR prefix .hic.hap2.p_ctg.gfa:
fully phased contig graph where each contig is fully phased.

.TP
*
.IR prefix .hic.a_ctg.gfa 
(optional with 
.BR --primary): 
assembly graph of alternate contigs.

.RE


.PP
Hifiasm keeps Hi-C alignment results and Hi-C index in two bin 
files: 
.B *hic.lk.bin
and 
.B *hic.tlb.bin.
Rerunning hifiasm with different Hi-C reads needs to delete these bin files
or use 
.BR -i .
.RE

.PP
Hifiasm generates the following assembly graphs only with HiFi reads:

.RS 2
.TP 2
*
.IR prefix .bp.p_ctg.gfa:
assembly graph of primary contigs. 

.TP
*
.IR prefix .bp.hap1.p_ctg.gfa:
partially phased contig graph of haplotype1.

.TP
*
.IR prefix .bp.hap2.p_ctg.gfa:
partially phased contig graph of haplotype2.
.RE

.PP
If the option
.BR -l0
or
.BR --primary
is specified, hifiasm outputs:

.RS 2
.TP 2
*
.IR prefix .p_ctg.gfa:
assembly graph of primary contigs.

.TP
*
.IR prefix .a_ctg.gfa:
assembly graph of alternate contigs.
.RE




.PP
For each graph, hifiasm also outputs a simplified version without sequences for
the ease of visualization. Hifiasm keeps corrected reads and overlaps in three
binary files such as it can regenerate assembly graphs from the binary files
without redoing error correction.