Deciphon is a robust protein annotation software tailored for handling long-read sequencing data with high error rates. By leveraging a novel approach that incorporates quasi-codons, Deciphon allows for direct annotation of protein sequences without the need for prior error correction or open reading frame identification.
The figure above illustrates the Pfam annotations found using Deciphon, compared to combinations of Prodigal and FragGeneScan with HMMER, on a long-read of the Chlamydia strain 14-2711_R47, isolated from flamingos. The alignment spans a chromosomal region of 23,953 nucleotides and consists of 23,242 matches, 880 deletions, and 711 insertions. Deciphon recovered 21 out of 29 proteins (72%) while mislabelling only one (Chordopox_A30L). FGS1-HMMER and FGS3-HMMER recovered 10 (34%) and 8 (28%) proteins, respectively. Prodigal-HMMER predicted 9 genes but found no significant match against Pfam.
To get started, you can install Deciphon via the command line using Python’s package manager:
pip install deciphonDeciphon is compatible with both macOS and Linux operating systems. Once installed, you can access the help documentation by entering:
deciphon --helpin the terminal to display usage information.
├─ c-core/ Deciphon core library written in C.
├─ cli/ Command-line interface (CLI) written in Python.
├─ compose/ Docker Compose setup for running a Deciphon server.
├─ control/ CLI for controlling and managing servers.
├─ intervals/ Python and R-based interval definitions.
├─ python-core/ Python wrapper for the Deciphon core library.
├─ sched/ RESTful API for Deciphon server.
├─ snap/ Python reader for Deciphon snap files.
If you find this project useful, please consider giving it a ⭐️!