Methods for computer processing sequence reads to detect molecular residual disease
Disclosed herein are methods for use in detection of molecular residual disease. The methods may comprise deep sequencing a panel of genomic regions in cell-free DNA molecules and computer processing sequence reads to detect variants that are indicative of molecular residual disease.
1. A method for detecting molecular residual disease in a subject, the method comprising:
(a) obtaining a plurality of cell-free deoxyribonucleic acid (cfDNA) molecules obtained or derived from the subject to generate amplified cfDNA molecules;
(b) enriching a plurality of the cfDNA molecules for a sequencing panel of no more than twenty genomic regions to generate an enriched set of cfDNA molecules, wherein the sequencing panel of no more than twenty genomic regions is selected using information derived from a cancer tumor biopsy of the subject indicating at least one of the no more than twenty genomic regions contain a tumor marker present in a greater proportion in the cancer tumor biopsy than a background population;
(c) sequencing a plurality of the enriched set of amplified cfDNA molecules at a sequence read depth of at least about fifty thousand sequence reads per base to generate sequence reads;
(d) computer processing a plurality of the sequence reads at least in part by aligning a plurality of the sequence reads to a reference genome to generate aligned sequence reads;
(e) computer processing a plurality of the aligned sequence reads to detect a variant corresponding to a tumor marker in the no more than twenty genomic regions indicated in step (b), thereby determining the molecular residual disease in the subject.
2. The method of claim 1 , wherein the sequencing is at a sequence read depth of at least about 100 thousand sequence reads per base.
3. The method of claim 1 , further comprising comparing sequence information obtained from the plurality of cfDNA molecules to sequence information obtained from a cohort of healthy individuals.
4. The method of claim 1 , wherein the computer processing in (e) comprises determining a plurality of base calls from the plurality of the aligned sequence reads at least in part by comparing the plurality of the aligned sequence reads with a plurality of reference sequences, wherein a base call is reported if its frequency is greater than a noise threshold as determined by the frequency in a plurality of reference sequences.
5. The method of claim 4 , wherein the reference sequences are sequences from healthy individuals.
6. The method of claim 1 , further comprising using a cohort of healthy individuals to provide a baseline frequency of a base at a position, and the baseline frequency is used in determining the plurality of base calls.
7. The method of claim 1 , further comprising using a frequency for a base at a position in a cohort of healthy individuals from the subject to a frequency of a base detected among sequence reads from the subject.
8. The method of claim 1 , wherein the enriched set of cfDNA molecules is representative of no more than about 35 thousand base pairs of a human genome.
9. The method of claim 1 , wherein the enriched set of cfDNA molecules is representative of no more than about 5 thousand base pairs of a human genome.
10. The method of claim 1 , wherein the plurality of cfDNA molecules is isolated from a bodily fluid sample.
11. The method of claim 10 , wherein the bodily fluid sample is a blood sample, a plasma sample, or a serum sample.
12. The method of claim 1 , wherein the subject has previously received a treatment for a cancer.
13. The method of claim 12 , wherein the cancer is selected from the group consisting of colorectal cancer, ovarian cancer, lung cancer, pancreatic cancer, and liver cancer.
14. The method of claim 12 , wherein the treatment for the cancer is selected from the group consisting of a surgical treatment, a radiation treatment, a chemotherapy, a targeted cancer therapeutics, and a cancer immunotherapy.
15. The method of claim 12 , wherein the subject does not detectably exhibit any symptoms of the cancer.
16. The method of claim 1 , further comprising determining a consensus sequence from a plurality of the sequence reads to reduce errors from the amplifying in (a) or the sequencing in (c).
17. The method of claim 1 , further comprising, prior to the sequencing in (c), tagging a plurality of the enriched set of cfDNA molecules by PCR amplification using primers with barcodes.
18. The method of claim 1 , wherein the sequencing in (c) is performed within a read budget that allocates a pre-determined total number of base reads, wherein the plurality of cfDNA molecules comprises no more than a pre-determined amount of DNA.
19. The method of claim 1 , wherein multiple cfDNA samples are collected from the subject over a plurality of time points and analyzed.
20. The method of claim 1 , wherein the sequencing panel of no more than twenty genomic regions is selected to achieve a sensitivity of at least 85% for a cancer selected from the group consisting of colorectal cancer, ovarian cancer, lung cancer, and pancreatic cancer.
21. The method of claim 1 , wherein the enriching in (b) is performed using amplification-based enrichment.
22. The method of claim 1 , further comprising filtering out at least a portion of the plurality of the sequence reads or the plurality of the aligned sequence reads.