sequence_align

KO | EN

gitlite — search

#hacktoberfest #react #ai #go #android #machine-learning #rust #macos #java #awesome #awesome-list #pytorch #security #openai

(★ 75)

Efficient implementations of Needleman-Wunsch and other sequence alignment algorithms written in Rust with Python bindings via PyO3.

sequence_align 최신버젼 다운로드

최종 버전 다운로드 (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# sequence_align
Efficient implementations of [Needleman-Wunsch](https://en.wikipedia.org/wiki/Needleman%E2%80%93Wunsch_algorithm)
and other sequence alignment algorithms written in Rust with Python bindings via [PyO3](https://github.com/PyO3/pyo3).
Supports both binary match/mismatch scoring and custom pairwise scoring functions for applications
like OCR text alignment, spatial matching, and other domains where continuous similarity measures
are needed.

## Installation
`sequence_align` is distributed via [PyPi](https://pypi.org/project/sequence_align) for Python 3.10 - 3.14, making installation as simple as the following --
no special setup required for cross-platform compatibility, Rust installation, etc.!

``` bash
pip install sequence_align
```

Alternatively, if one wishes to develop for `sequence_align`, first ensure that both
[Python](https://wiki.python.org/moin/BeginnersGuide/Download) and [Rust](https://www.rust-lang.org/tools/install)
are installed on your system. Then, install [Maturin](https://www.maturin.rs/#usage) and run
`maturin develop` (optionally with the `-r` flag to compile a release build, instead of an unoptimized debug build)
from the root of your cloned repo to build and install `sequence_align` in your active Python environment.

## Quick Start
Pairwise sequence algorithms are available in [`sequence_align.pairwise`](src/sequence_align/pairwise.py).
The following algorithms are implemented:

- [**Needleman-Wunsch**](https://en.wikipedia.org/wiki/Needleman%E2%80%93Wunsch_algorithm): Global sequence alignment with `O(M*N)` time and space.
- [**Needleman-Wunsch with custom scores**](https://en.wikipedia.org/wiki/Needleman%E2%80%93Wunsch_algorithm): A variant that accepts a custom pairwise scoring function `score_fn(a, b) -> float` instead of flat match/mismatch scores. This is useful when alignment quality depends on continuous similarity measures rather than binary element equality.
- [**Hirschberg**](https://en.wikipedia.org/wiki/Hirschberg%27s_algorithm): A modification of Needleman-Wunsch with the same `O(M*N)` time complexity but only `O(min{M, N})` space, making it an appealing option for memory-limited applications or extremely large sequences.

One may also compute the Needleman-Wunsch alignment score for alignments produced by any of the above algorithms
using [`alignment_score`](src/sequence_align/pairwise.py).

Using these algorithms is straightforward:

``` python
from sequence_align.pairwise import (
    alignment_score,
    hirschberg,
    needleman_wunsch,
    needleman_wunsch_with_scores,
)

# See https://en.wikipedia.org/wiki/Needleman%E2%80%93Wunsch_algorithm#/media/File:Needleman-Wunsch_pairwise_sequence_alignment.png
# Use Needleman-Wunsch default scores (match=1, mismatch=-1, indel=-1)
seq_a = ["G", "A", "T", "T", "A", "C", "A"]
seq_b = ["G", "C", "A", "T", "G", "C", "G"]

aligned_seq_a, aligned_seq_b = needleman_wunsch(
    seq_a,
    seq_b,
    "_",  # Represent gaps with this value
    match_score=1.0,
    mismatch_score=-1.0,
    indel_score=-1.0,
)

# Expects ["G", "_", "A", "T", "T", "A", "C", "A"]
print(aligned_seq_a)

# Expects ["G", "C", "A", "_", "T", "G", "C", "G"]
print(aligned_seq_b)

# Expects 0
score = alignment_score(
    aligned_seq_a,
    aligned_seq_b,
    "_",
    match_score=1.0,
    mismatch_score=-1.0,
    indel_score=-1.0,
)
print(score)

# See https://en.wikipedia.org/wiki/Hirschberg%27s_algorithm#Example
seq_a = ["A", "G", "T", "A", "C", "G", "C", "A"]
seq_b = ["T", "A", "T", "G", "C"]

aligned_seq_a, aligned_seq_b = hirschberg(
    seq_a,
    seq_b,
    "_",
    match_score=2.0,
    mismatch_score=-1.0,
    indel_score=-2.0,
)

# Expects ["A", "G", "T", "A", "C", "G", "C", "A"]
print(aligned_seq_a)

# Expects ["_", "_", "T", "A", "T", "G", "C", "_"]
print(aligned_seq_b)

# Expects 1
score = alignment_score(
    aligned_seq_a,
    aligned_seq_b,
    "_",
    match_score=2.0,
    mismatch_score=-1.0,
    indel_score=-2.0,
)
print(score)

# Custom pairwise scoring: align words using character overlap similarity
words_a = ["hello", "world", "foo"]
words_b = ["hallo", "welt", "baz", "foo"]

def char_overlap_score(a: str, b: str) -> float:
    """Score based on character-level overlap between two words."""
    if a == b:
        return 2.0
    shared = len(set(a) & set(b))
    total = len(set(a) | set(b))
    return (2.0 * shared / total) - 1.0 if total > 0 else -1.0

aligned_words_a, aligned_words_b = needleman_wunsch_with_scores(
    words_a,
    words_b,
    "_",
    score_fn=char_overlap_score,
    indel_score=-1.0,
)

# Expects ["hello", "world", "_", "foo"]
print(aligned_words_a)

# Expects ["hallo", "welt", "baz", "foo"]
print(aligned_words_b)
```

## Development

To set up a local development environment, ensure that both
[Python](https://wiki.python.org/moin/BeginnersGuide/Download) and [Rust](https://www.rust-lang.org/tools/install)
are installed, then:

``` bash
maturin develop -r  # build and install in the active Python environment
./scripts/test.sh   # run tests via pytest
./scripts/lint.sh   # run all linters (ruff, mypy, cargo fmt, cargo clippy)
./scripts/lint.sh --fix  # auto-fix where possible
```

## Performance Benchmarks
All tests below were conducted sequentially on a [AWS R5.4 instance](https://aws.amazon.com/ec2/instance-types/r5/)
with 16 cores and 128 GB of memory. The pair of sequences for alignment consist of a character sequence of randomly
selected A/C/G/T nucleotide bases along with another that is identical, except with 10% of the characters randomly
perturbed by deletion, insertion of another randomly-selected character after the entry, or replacement with a
different randomly-selected character.

As one can see, while `sequence_align` is comparable to some other toolkits in terms of speed, its memory performance
is **best-in-class**, even when compared to toolkits using the same algorithm, such as Needleman-Wunsch being used in
`pyseq-align`.

_(Please note that some lines terminate early, as some toolkits took prohibitively long and/or ran out of memory at higher scales.)_

## Changelog
See [CHANGELOG.md](CHANGELOG.md) for a full list of changes across versions.

## License
Licensed under the Apache 2.0 License. Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

sequence_align

sequence_align 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

sequence_align

sequence_align 최신버젼 다운로드

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README