Gene Sequencing with Artificial Intelligence and Nuclear Magnetic Resonance (NMR) Spectroscopy

Artificial intelligence (AI) can generate gene sequences using DNA samples from organisms. This is achieved through Artificial Neural Networks (ANNs), particularly Convolutional Neural Networks (CNNs), which excel in image classification tasks. Specifically, a classifier CNN can be trained on a dataset composed of NMR spectra of nitrogenous bases.To initiate this process, the DNA sample undergoes NMR spectroscopy to produce molecular structures. The classifier CNN is then trained using data representing the molecular structures of the four nitrogenous bases: Adenine, Guanine, Thymine, and Cytosine. During training, the CNN identifies these molecular structures within the sequence and labels them as 'A', 'G', 'T', or 'C' accordingly.Upon receiving the NMR-derived molecular structures of these bases as inputs sequentially, the model can classify them into their respective genetic letters ('A', 'G', 'T', 'C'), thereby generating the gene sequence. Read more ...Garg, R. , Vyas, A. , Khan, A. , Tariq, M. (2024), 'Codes beyond Bits and Bytes: A Blueprint for Artificial Life', World Academy of Science, Engineering and Technology, Open Science Index 213, International Journal of Biotechnology and Bioengineering, 18(9), 114 - 126.人工智能（Artificial Intelligence）可利用生物体的脱氧核糖核酸（DNA）样本生成基因序列。该技术依托人工神经网络（Artificial Neural Networks，ANNs）实现，其中尤以卷积神经网络（Convolutional Neural Networks，CNNs）表现突出——这类模型在图像分类任务中具备卓越性能。具体而言，可基于含氮碱基的核磁共振（Nuclear Magnetic Resonance，NMR）光谱数据集训练分类器卷积神经网络。启动该流程前，需先对DNA样本开展核磁共振光谱分析以获取其分子结构。随后，利用四种含氮碱基：腺嘌呤（Adenine）、鸟嘌呤（Guanine）、胸腺嘧啶（Thymine）与胞嘧啶（Cytosine）的分子结构数据，对该分类器卷积神经网络进行训练。训练过程中，卷积神经网络会识别序列中的各类分子结构，并依此将其标记为'A'、'G'、'T'或'C'。当模型依次接收经核磁共振光谱解析得到的碱基分子结构作为输入时，即可将其分类至对应的遗传字母（'A'、'G'、'T'、'C'），进而完成基因序列的生成。更多内容详见……Garg R、Vyas A、Khan A、Tariq M. (2024)，《超越比特与字节的编码：人工生命蓝图》，世界科学、工程与技术学院，开放科学索引213，《国际生物技术与生物工程期刊》，18(9)，114-126。