/ Framework

A strongly-typed and statically-compiled high-performance Pythonic language

A strongly-typed and statically-compiled high-performance Pythonic language

Seq — a language for bioinformatics

Seq is a programming language for computational genomics and bioinformatics. With a Python-compatible syntax and a host of domain-specific features and optimizations, Seq makes writing high-performance genomics software as easy as writing Python code, and achieves performance comparable to (and in many cases better than) C/C++.

Think of Seq as a strongly-typed and statically-compiled Python: all the bells and whistles of Python, boosted with strong type system, without any performance overhead.

Seq is able to outperform Python code by up to 160x. Seq can further beat equivalent C/C++ code by up to 2x without any manual interventions, and also natively supports parallelism out of the box. Implementation details and benchmarks are discussed in our paper.

Example

Seq is a Python-compatible language, and the vast majority of Python programs should work without any modifications:

def check_prime(n):
    if n > 1:
        for i in range(2, n):
            if n % i == 0:
                return False
        return True
    else:
        return False

n = 1009
print n, 'is', 'a' if check_prime(n) else 'not a', 'prime'

Here is an example that showcases Seq's bioinformatics features: a seeding application in Seq using a hypothetical genome index, like what is typically found in seed-and-extend alignment algorithms:

from sys import argv
from genomeindex import *
type K = Kmer[20]

# index and process 20-mers
def process(kmer: K,
            index: GenomeIndex[K]):
    prefetch index[kmer], index[~kmer]
    hits_fwd = index[kmer]
    hits_rev = index[~kmer]
    ...

# index over 20-mers
index = GenomeIndex[K](argv[1])

# stride for k-merization
stride = 10

# sequence-processing pipeline
(FASTQ(argv[2])
  |> seqs
  |> kmers[K](stride)
  |> process(index))

A few notable aspects of this code:

  • Seq provides native k-mer types, e.g. a 20-mer is represented by Kmer[20] as above.
  • k-mers can be reverse-complemented with ~.
  • Seq provides easy iteration over common formats like FASTQ (FASTQ above).
  • Complex pipelines are easily expressible in Seq (via the |> syntax).
  • Seq can perform pipeline transformations to make genomic index lookups faster via prefetch.

For a concrete example of genomeindex, check out our re-implementation of SNAP's index.

Install

Pre-built binaries

Pre-built binaries for Linux and macOS on x86_64 are available alongside each release. We also have a script for downloading and installing pre-built versions:

wget -O - https://raw.githubusercontent.com/seq-lang/seq/master/install.sh | bash

This will install Seq in a new .seq directory within your home directory. Be sure to update ~/.bash_profile as the script indicates afterwards!

Seq binaries require a libomp to be present on your machine. brew install libomp or apt install libomp5 should do the trick.

Documentation

Please check seq-lang.org for in-depth documentation.

Citing Seq

If you use Seq in your research, please cite:

Ariya Shajii, Ibrahim Numanagić, Riyadh Baghdadi, Bonnie Berger, and Saman Amarasinghe. 2019. Seq: a high-performance language for bioinformatics. Proc. ACM Program. Lang. 3, OOPSLA, Article 125 (October 2019), 29 pages. DOI: https://doi.org/10.1145/3360551

BibTeX:

@article{Shajii:2019:SHL:3366395.3360551,
 author = {Shajii, Ariya and Numanagi\'{c}, Ibrahim and Baghdadi, Riyadh and Berger, Bonnie and Amarasinghe, Saman},
 title = {Seq: A High-performance Language for Bioinformatics},
 journal = {Proc. ACM Program. Lang.},
 issue_date = {October 2019},
 volume = {3},
 number = {OOPSLA},
 month = oct,
 year = {2019},
 issn = {2475-1421},
 pages = {125:1--125:29},
 articleno = {125},
 numpages = {29},
 url = {http://doi.acm.org/10.1145/3360551},
 doi = {10.1145/3360551},
 acmid = {3360551},
 publisher = {ACM},
 address = {New York, NY, USA},
 keywords = {Python, bioinformatics, computational biology, domain-specific language, optimization, programming language},
}

GitHub

Comments