daniel-liu-c0deb0t / antisequence

Rust library for processing sequencing reads.

ANTISEQUENCE

Rust library for processing sequencing reads.

Work in progress! Very early stages of the project.

Goals

• Robust, flexible, and actually universal primitives for manipulating raw DNA/RNA sequences from fastq files
• Blazing fast and scalable implementation using SIMD and CPU parallelization
• Simple interface allowing anyone (even non-Rustaceans) to pick up and use
• Extensible with custom Rust code and embeddable into existing pipelines

ANTISEQUENCE should enable you to build robust, efficient, and production-ready pipeline for your custom sequencing data.

K-pop song

Enjoy a K-pop song.

antisequence's Issues

Better expression API for quality scores

Currently, whenever quality scores are required, the same expression is evaluated but with all strings substituted with quality scores. This actually works as expected for (just) rearrangements, slice, concat, and even revcomp. However, there are many (perhaps uncommon) edge cases where this behavior is not desired. On the other hand, it is more tedious to require writing separate quality-score-specific expressions. So is there a better API?

Ideas for optimization passes

Separate passes will be needed for expressions and operations.

Expressions

• constant propagation

Operations

• combine trims

More ideas are welcome!

Broadcast operations

Many operations, like MatchAny, could output multiple labels (a list of labels) to indicate multiple matches. Then, it would be useful for all operations to support broadcasting their operation across a list of labels.

Demo using codespaces and perfetto

Create an example project on codespaces to demo ANTISEQUENCE.

To make debugging easier, output timings for each operation in JSON format (gated by boolean parameter), which can be visualized in Perfetto. Also output the reads after each operation.