The invention discloses a
chromosome outer
circular DNA sequence identification method based on
deep learning, and relates to the technical field of
genome biological information, and the method comprises the following steps: obtaining a known
chromosome outer
circular DNA sequence as a
positive sample; randomly selecting a
genome segment as a
negative sample, wherein the overlapping degree of the
genome segment and any known
chromosome outer
circular DNA region is lower than 20% from the
reference genome according to an equal
length distribution and chromosome distribution principle; cleaning and standardizing the positive and negative samples, and dividing the positive and negative samples into a
training set, a
verification set and a
test set; converting the positive and
negative sample sequences into feature representation which can be processed by a neural network through a multi-
modal coding strategy, and training a one-dimensional
convolutional neural network architecture by using a
training set; when the number of the training samples is insufficient, introducing a deep convolutional
generative adversarial network to perform data enhancement; and
processing a DNA sequence to be detected through a multi-
modal coding strategy, inputting the processed
DNA sequence to the trained model, outputting a probability
score of existence of the circular
DNA outside the chromosome, and completing classification judgment according to a preset threshold value.