Shasta long read assembler
This is the documentation for the
Shasta long read assembler.
If you are seeing this documentation on github.io,
it applies to the latest version of Shasta
on GitHub (not necessarily the same as the latest release).
Documentation for any version of Shasta is available
in the source code in the shasta/docs directory
and in any build under the shasta-build/shasta-install/docs
directory.
Abstract
The goal of Shasta is to rapidly produce accurate assembled sequence using as input DNA reads generated by Oxford Nanopore flow cells.
Computational methods used by the Shasta assembler include:
- Using a run-length representation of the read sequence. This makes the assembly process more resilient to errors in homopolymer repeat counts, which are the most common type of errors in Oxford Nanopore reads.
- Using in some phases of the computation a representation of the read sequence based on markers, a fixed subset of short k-mers (k ≈ 10).
An current implementation of Shasta is complete and functional,
but significant improvements in several areas are possible.
As implemented, it can run an assembly of a human genome at coverage around 60x
in about 5 hours using a single, large machine (AWS instance type
x1.32xlarge, with 128 virtual processors and 1952 GB of memory).
The compute cost of such an assembly is around $20 at AWS spot market or reserved prices.
See Shafin et al, Nature Biotechnology 2020 for an error analysis of the Shasta assembler and more.