rust-mdbg: Minimizer-space de Bruijn graphs (mdBG) for whole-genome assembly
rust-mdbg is an ultra-fast minimizer-space de Bruijn graph (mdBG)
implementation, geared towards the assembly of long and accurate reads such as
PacBio HiFi.
Rationale
rust-mdbg performs mdBG construction of a 52x human genome HiFi data in around
10 minutes on 8 threads, with 10GB of maximum RAM usage.
rust-mdbg is fast because it operates in minimizer-space, meaning that the
reads, the assembly graph, and the final assembly, are all represented as
ordered lists of minimizers, instead of strings of nucleotides. A conversion
step then yields a classical base-space representation.
Limitations
However, this high speed comes at a cost! :)
- rust-mdbg gives good-quality results but still of lower contiguity and
completeness than state-of-the-art assemblers such as HiCanu and hifiasm.
- rust-mdbg performs best with at least 40x to 50x of coverage.
- No polishing step is implemented; so, assemblies will have around the same
accuracy as the reads.
- Cannot assemble Nanopore data due to its higher error rate (see this comment)
Location and version:
1
2
3
4
|
$ which rust-mdbg
/local/cluster/bin/rust-mdbg
$ rust-mdbg --version
rust-mdbg 0.1.0
|
help message:
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
|
$ rust-mdbg --help
rust-mdbg 0.1.0
Original implementation of minimizer-space de Bruijn graphs (mdBG) for genome assembly.
rust-mdbg is an ultra-fast minimizer-space de Bruijn graph (mdBG) implementation, geared towards the assembly of long
and accurate reads such as PacBio HiFi. rust-mdbg is fast because it operates in minimizer-space, meaning that the
reads, the assembly graph, and the final assembly, are all represented as ordered lists of minimizers, instead of
strings of nucleotides. A conversion step then yields a classical base-space representation.
USAGE:
rust-mdbg [FLAGS] [OPTIONS] [reads]
FLAGS:
--bf
Enable Bloom filters
Bloom filters can be used to reduce memory usage, but results in slightly less contiguous assemblies.
--debug
Activate debug mode
Debug mode shows the base-space sequence and the minimizer-space representationobtained from each read.
--error-correct
Enable error correction with minimizer-space POA
Partial order alignment (POA) can be used (in minimizer-space) to error-correctreads with up to 5% error
rate.
-h, --help
Prints help information
--hpc
Homopolymer-compressed (HPC) input
Both raw and homopolymer-compressed (HPC) reads can be provided as input. If the reads are not compressed,
rust-mdbg manually performs HPC, but uses the raw sequences for transformation into base-space.
--reference
Reference genome input
Indicates that the input is a (single or a set of) genome(s), not reads. Allowsmulti-line FASTA and doesn't
filter any kminmers
--restart-from-postcor
Assemble error-corrected reads
Reads that are error-corrected with rust-mdbg can be used as input.
-V, --version
Prints version information
OPTIONS:
--correction-threshold <correction-threshold>
POA correction threshold
The maximum number of reads in a minimizer-space partial order alignment (POA) graph that can be replaced
with the consensus.
-d, --density <density>
Density threshold for density-based selection scheme
The density threshold is analogous to the fraction of l-mers that will be selected as minimizers from a
read.
--distance <distance>
Distance metric (0: Jaccard, 1: containment, 2: Mash)
rust-mdbg uses a distance metric to filter out reads that are not very similar to a query read before
constructing a partial order alignment (POA) graph.
-k, --k <k>
k-min-mer length
The length of each node of the mdBG. If fewer l-mers than this value are obtained from a read, they will be
ignored.
-l, --l <l>
l-mer (minimizer) length
The length of each minimizer selected using the minimizer scheme from base-space sequences.
--lcp <lcp>
Core substring file (enables locally consistent parsing (LCP))
Core substrings need to be provided as a single file, one core substring per line. The minimizers will be
selected if they are a core substring and if they satisfy the density bound.
--lmer-counts <lmer-counts>
l-mer counts (enables downweighting of frequent l-mers)
Frequencies of l-mers in the reads (obtained using k-mer counters) can be provided in order to downweight
frequently-occurring l-mers and increase contiguity.
--lmer-counts-max <lmer-counts-max>
Maximum l-mer count threshold
l-mers with frequencies above this threshold will be downweighted.
--lmer-counts-min <lmer-counts-min>
Minimum l-mer count threshold
l-mers with frequencies below this threshold will be downweighted.
--minabund <minabund>
Minimum k-min-mer abundance
k-min-mers that occur fewer times than this value will be removed from the mdBG.
-n, --n <n>
Tuple length for bucketing similar reads
Reads that share a tuple (of l-mers) of this length will be bucketed as candidates for their respective
partial order alignment (POA) graphs.
-p, --prefix <prefix>
Output prefix for GFA and .sequences files
All files generated by rust-mdbg (including the GFA and .sequences output files) will have this file name.
--presimp <presimp>
Pre-simplification (pre-simp) threshold
Additional graph simplification heuristics prior to the construction of the mdBG can be performed. The
default value for the pre-simplification step is 0.1.
-t, --t <t>
POA path weight threshold
During the consensus step for a partial order alignment (POA) graph, paths withweights below this value
will be removed.
--threads <threads>
Number of threads
rust-mdbg is highly parallelized to decrease running time, but can be run on a single core as well.
--uhs <uhs>
Universal k-mer file (enables universal hitting sets (UHS))
Universal k-mers need to be provided as a single file, one universal k-mer per line. The minimizers will be
selected if they are a universal k-mer and if they satisfy the density bound.
ARGS:
<reads>
Input file (raw or gzip-/lz4-compressed FASTX)
Input file can be FASTA/FASTQ, as well as gzip-compressed (.gz) or lz4-compressed (.lz4). Lowercase bases
are currently not supported; see documentation for formatting.
|
Associated utilities in your $PATH:
1
2
3
|
magic_simplify
multik
to_basespace
|
Also, gfatools is required, which is also installed.
software ref: https://github.com/ekimb/rust-mdbg
research ref: https://doi.org/10.1016/j.cels.2021.08.009