Transcription factor profiling
Methods and systems disclosed herein can improve analysis capabilities of genomic materials. The methods provided herein may examine transcription factor binding site accessibility to diagnose a disease or monitor progression of a disease in a subject.
1 . A method for treating a cancer in a human subject, said method comprising:
(a) providing a set of sequence reads obtained by whole genome sequencing of cell-free deoxyribonucleic acid (cfDNA) obtained or derived from said human subject;
(b) aligning each sequence read of said set of sequence reads to a human reference genome to provide a set of aligned sequence reads;
(c) for each binding site of a set of binding sites of a transcription factor, selecting a genomic region of said set of aligned sequence reads that flanks each binding site of said set of binding sites to provide a set of flanking genomic regions, wherein each binding site of said set of binding sites exhibits differential accessibility to a target enhancer of said transcription factor in a first population of human subjects having said cancer as compared to a second population of human subjects not having said cancer, wherein said cancer comprises breast cancer, colon cancer, prostate adenocarcinoma, or small-cell neuroendocrine prostate cancer, wherein said transcription factor is selected from the group consisting of GRHL1, GRHL2, GRHL3, AR, ASH-2, HOXB13, EVX2, PU.1, LYL1, SPIB, FOXA1, FOXA2, GATA2, GATA3, ZNF121, HNF1A, HNF4A, HNF4G, DLX2, REST, GLIS1, SOX2, SOX11, and NKX3-1, and wherein each binding site of said set of binding sites is different;
(d) for each flanking genomic region of said set of flanking genomic regions, using at least a portion of said set of aligned sequence reads to generate a read depth coverage pattern, thereby providing a set of read depth coverage patterns for said set of binding sites of said transcription factor, respectively, wherein said set of read depth coverage patterns comprise nucleosome data;
(e) processing said set of read depth coverage patterns to provide a signal, wherein said signal characterizes binding site accessibility of said set of binding sites of said transcription factor;
(f) splitting said set of read depth coverage patterns into a low-frequency signal and a high-frequency signal using a low-pass filter, wherein said high-frequency signal is a measure of binding site accessibility for a given binding site, and wherein said splitting comprises applying a detrending filter to suppress effects on said read depth coverage pattern not contributed by preferential nucleosome positioning and to remove local biases from said nucleosome data;
(g) determining an accessibility score for each binding site of said set of binding sites of said transcription factor, wherein said determining comprises using said low-frequency signal to normalize said high-frequency signal;
(h) detecting a presence of said cancer in said human subject, based at least in part on a deviation of said accessibility score from a reference accessibility score; and
(i) upon detecting said presence of said cancer in said human subject, administering a therapeutic intervention to said human subject, thereby treating said cancer.
2 . The method of claim 1 , wherein said accessibility score that deviates from said reference accessibility score by greater than ±3 mean of a standard deviation of a set of accessibility scores for each binding site of said set of binding sites of said transcription factor, when said accessibility score is calculated using a Savitzky-Golay filter, is indicative of said presence of said cancer in said human subject.
3 . The method of claim 2 , wherein said accessibility score comprises a z-score determined based on said standard deviation of said set of accessibility scores for each binding site of said set of binding sites of said transcription factor.
4 . The method of claim 1 , wherein said transcription factor is selected from the group consisting of FOXA1, GRHL1, GRHL2, and GRHL3.
5 . The method of claim 1 , wherein said transcription factor is selected from the group consisting of GRHL2, PU.1, LYL1, SPIB, AR, HOXB13, and NKX3-1.
6 . The method of claim 1 , wherein (h) further comprises identifying an increased accessibility score of a binding site of said transcription factor as compared to said reference accessibility score.
7 . The method of claim 1 , wherein (h) further comprises applying a trained machine learning algorithm to said accessibility score for each of said binding sites of said set of binding sites of said transcription factor to detect said presence of said cancer in said human subject.
8 . The method of claim 7 , wherein said trained machine learning algorithm is selected from the group consisting of a regression, a support vector machine, a tree-based method, a neural network, and a random forest.
9 . The method of claim 1 , wherein (g) further comprises performing a locally weighted smoothing using said high-frequency signal.
10 . The method of claim 9 , wherein said locally weighted smoothing comprises a locally estimated scatterplot smoothing (LOESS).
11 . The method of claim 1 , wherein said set of binding sites of said transcription factor comprises at least 25 different binding sites of said transcription factor.
12 . The method of claim 1 , wherein said set of sequence reads comprises at least 50 million sequence reads.
13 . The method of claim 1 , wherein said transcription factor is selected from the group consisting of GRHL2, FOXA1, and ZNF121; and wherein said cancer comprises said breast cancer.
14 . The method of claim 1 , wherein said transcription factor is selected from the group consisting of EVX2, DLX2, HNF1A, HNF4A, GRHL2, and HNF4G; and wherein said cancer comprises said colon cancer.
15 . The method of claim 1 , wherein said transcription factor is selected from the group consisting of LYL1 and PU.1; and wherein said cancer comprises said colon cancer.
16 . The method of claim 1 , wherein said transcription factor is selected from the group consisting of GRHL2, FOXA1, HOXB13, AR, and NKX3-1; and wherein said cancer comprises said prostate adenocarcinoma.
17 . The method of claim 1 , wherein said transcription factor is selected from the group consisting of REST, GRHL2, GRHL3, FOXA1, FOXA2, GATA2, GATA3, HOXB13, AR, and NKX3-1; and wherein said cancer comprises said prostate small-cell neuroendocrine prostate cancer.
18 . The method of claim 1 , wherein said transcription factor is selected from the group consisting of GLIS1, SOX2, and SOX11; and wherein said cancer comprises said small-cell neuroendocrine prostate cancer.
19 . The method of claim 1 , wherein said low-pass filter comprises a Savitzky-Golay filter.
20 . The method of claim 19 , wherein said Savitzky-Golay filter comprises a third-order Savitzky-Golay filter.
21 . The method of claim 19 , further comprising obtaining said low-frequency signal using a first Savitzky-Golay filter with a first window size, and obtaining said high-frequency signal using a second Savitzky-Golay filter with a second window size, wherein said first window size is larger than said second window size.
22 . The method of claim 1 , further comprising performing whole genome sequencing of said cfDNA obtained or derived from said human subject, thereby generating said set of sequence reads.
23 . The method of claim 22 , further comprising obtaining or deriving said cfDNA from a biological sample of said human subject.
24 . The method of claim 23 , wherein said biological sample is a blood sample, a plasma sample, a serum sample, a sweat sample, a urine sample, a saliva sample, a cell sample, a tissue sample, or a derivative thereof.
25 . The method of claim 24 , wherein said biological sample is said plasma sample.
26 . The method of claim 1 , further comprising, responsive at least in part to said detecting in (h), performing on said human subject a secondary clinical test, wherein said secondary clinical test comprises an imaging test, a blood test, a computed tomography (CT) scan, a magnetic resonance imaging (MRI) scan, an ultrasound scan, a chest X-ray, a positron emission tomography (PET) scan, a PET-CT scan, a cytology assay, or a combination thereof.
27 . The method of claim 1 , wherein said human reference genome is GrCH38, GrCH37, hg19, or hg38.
28 . The method of claim 1 , wherein said transcription factor is selected from the group consisting of ASH-2, HOX-B13, EVX2, PU.1, Lyl-1, Spi-B, FOXA1, HNF-1A, HNF-4A, HNF-4G, and DLX-2.
29 . The method of claim 28 , wherein said transcription factor comprises GRHL2.
30 . The method of claim 1 , wherein said detrending filter comprises a Savitzky-Golay filter.
31 . The method of claim 1 , wherein said therapeutic intervention comprises a chemotherapy, a surgery, or an androgen deprivation therapy.
32 . The method of claim 31 , wherein said therapeutic intervention comprises said chemotherapy.
33 . The method of claim 31 , wherein said therapeutic intervention comprises said surgery.
34 . The method of claim 31 , wherein said therapeutic intervention comprises said androgen deprivation therapy.