A Deep-Learning Pipeline for TSS Coverage Imputation From Shallow Cell-Free DNA Sequencing.

Cell-free DNA (cfDNA) serves as a footprint of the nucleosome occupancy status of transcription start sites (TSSs), and has been subject to wide development for use in noninvasive health monitoring and disease detection. However, the requirement for high sequencing depth limits its clinical use. Here, we introduce a deep-learning pipeline designed for TSS coverage profiles generated from shallow cfDNA sequencing called the Autoencoder of cfDNA TSS (AECT) coverage profile. AECT outperformed existing single-cell sequencing imputation algorithms in terms of improvements to TSS coverage accuracy and the capture of latent biological features that distinguish sex or tumor status. We built classifiers for the detection of breast and rectal cancer using AECT-imputed shallow sequencing data, and their performance was close to that achieved by high-depth sequencing, suggesting that AECT could provide a broadly applicable noninvasive screening approach with high accuracy and at a moderate cost.

Noninvasive discrimination of benign and malignant breast lesions using genome-wide nucleosome profiles of plasma cell-free DNA.

BACKGROUND: Breast malignancy is the most frequently diagnosed malignancy in women worldwide, and the diagnosis relies on invasive examinations. However, most clinical breast changes in women are benign, and invasive diagnostic approaches cause unnecessary suffering for the patients. Thus, a novel noninvasive approach for discriminating malignant breast lesions from benign lesions is needed. METHODS: We performed cell-free DNA (cfDNA) sequencing on plasma samples from 173 malignant breast lesion patients, 158 benign breast lesion patients, and 102 healthy women. We then analyzed the cfDNA-based nucleosome profiles, which reflect the various tissues of origin and transcription factor activities. Moreover, by using machine learning classifiers along with the cfDNA sequencing data, we built classifiers for discriminating benign from malignant breast lesions. Receiver operating characteristic curve analyses were used to evaluate the performance of the classifiers. RESULTS: cfDNA-based nucleosome profiles reflected the various tissues of origin and transcription factor activities in benign and malignant breast lesions. The cfDNA-based transcription factor activities and breast malignancy-specific transcription factor-binding site accessibility profiles could accurately distinguish benign and malignant breast lesions, with area under the curve values of 0.777 and 0.824, respectively. CONCLUSIONS: Our proof-of-principle study established a methodology for noninvasively discriminating benign from malignant breast lesions.

Noninvasive inferring expressed genes and in vivo monitoring of the physiology and pathology of pregnancy using cell-free DNA.

BACKGROUND: Noninvasive monitoring of fetal development and the early detection of pregnancy-associated complications is challenging, largely because of the lack of information about the molecular spectrum during pregnancy. Recently, cell-free DNA in plasma was found to reflect the global nucleosome footprint and status of gene expression and showed potential for noninvasive health monitoring during pregnancy. OBJECTIVE: We aimed to test the relationships between plasma cell-free DNA profiles and pregnancy biology and evaluate the use of a cell-free DNA profile as a noninvasive method for physiological and pathologic status monitoring during pregnancy. STUDY DESIGN: We used genome cell-free DNA sequencing data generated from noninvasive prenatal testing in a total of 2937 pregnant women. For each physiological and pathologic condition, features of the cell-free DNA profile were identified using the discovery cohort, and support vector machine classifiers were built and evaluated using independent training and validation cohorts. RESULTS: We established nucleosome occupancy profiles at transcription start sites in different gestational trimesters, demonstrated the relationships between gene expression and cell-free DNA coverage at transcription start sites, and showed that the cell-free DNA profiles at transcription start sites represented the biological processes of pregnancy. In addition, using cell-free DNA data, nucleosome profiles of transcription factor binding sites were identified to reflect the transcription factor footprint, which may help to reveal the molecular mechanisms underlying pregnancy. Finally, by using machine-learning models on low-coverage noninvasive prenatal testing data, we evaluated the use of cell-free DNA nucleosome profiles for distinguishing gestational trimesters, fetal sex, and fetal trisomy 21 and highlighted its potential utility for predicting physiological and pathologic fetal conditions by using low-coverage noninvasive prenatal testing data. CONCLUSION: Our analyses profiled nucleosome footprints and regulatory networks during pregnancy and established a noninvasive proof-of-principle methodology for health monitoring during pregnancy.

Simultaneous detection of target CNVs and SNVs of thalassemia by multiplex PCR and next­generation sequencing.

Thalassemia is caused by complex mechanisms, including copy number variants (CNVs) and single nucleotide variants (SNVs). The CNV types of α­thalassemia are typically detected by gap­polymerase chain reaction (PCR). The SNV types are detected by Sanger sequencing. In the present study, a novel method was developed that simultaneously detects CNVs and SNVs by multiplex PCR and next­generation sequencing (NGS). To detect CNVs, 33 normal samples were used as a cluster of control values to build a baseline, and the A, B, C, and D ratios were developed to evaluate­SEA, ­α4.2, ­α3.7, and compound or homozygous CNVs, respectively. To detect other SNVs, sequencing data were analyzed using the system's software and annotated using Annovar software. In a test of performance, 128 patients with thalassemia were detected using the method developed and were confirmed by Sanger sequencing and gap­PCR. Four different CNV types were clearly distinguished by the developed algorithm, with ­SEA, ­α3.7, ­α4.2, and compound or homozygous deletions. The sensitivities for each CNV type were 96.72% (59/61), 97.37% (37/38), 83.33% (10/12) and 95% (19/20), and the specificities were 93.94% (32/33), 93.94% (32/33), 100% (33/33) and 100% (33/33), respectively. The SNVs detected were consistent with those of the Sanger sequencing.