sequmap 0.1.0

An AA/NT sequence encoder using umap & pca

sequmap

Encode and embed sequences as kmer vectors (both NT and AA) and generate projections using UMAP

Setup

pip install sequmap

Usage

Embed and project sequences using

from sequmap import sequmap, NT_DICT
proj = sequmap(seqs, ndim, k = 5, lookup_dic = NT_DICT, n_neighbors = 30, min_dist = 0.001) 

where seqs is an array of strings, ndim is the number of projected dimensions and k is the kmer size. By default, sequences are embedded by counting all unique kmers and the distance between sequences is estimated using the SquaredEuclidean distance between dimensionality-reduced kmer vector.

Additionally, a simpler seqpca method is included which just performs PCA in the kmer domain, used as follows:

from sequmap import seqpca, NT_DICT
proj = seqpca(seqs, ndim, k = 5, lookup_dic = NT_DICT) 

The following lookup dictionaries are included for sequence encoding:

AA_DICT = Dict(
    'A' => 1, 'C' => 2, 'D' => 3, 'E' => 4, 'F' => 5, 
    'G' => 6, 'H' => 7, 'I' => 8, 'K' => 9, 'L' => 10, 
    'M' => 11, 'N' => 12, 'P' => 13, 'Q' => 14, 'R' => 15, 
    'S' => 16, 'T' => 17, 'V' => 18, 'W' => 19, 'Y' => 20, 
    '*' => 21, 'X' => 22
    )

NT_DICT = Dict('A' => 1, 'C' => 2, 'G' => 3, 'T' => 4)