Distinguishing methylation levels in complex biological samples

Abstract

Provided herein is a method for distinguishing an aberrant methylation level for DNA from a first cell type, including steps of (a) providing a test data set that includes (i) methylation states for a plurality of sites from test genomic DNA from at least one test organism, and (ii) coverage at each of the sites for detection of the methylation states; (b) providing methylation states for the plurality of sites in reference genomic DNA from one or more reference individual organisms, (c) determining, for each of the sites, the methylation difference between the test genomic DNA and the reference genomic DNA, thereby providing a normalized methylation difference for each site; and (d) weighting the normalized methylation difference for each site by the coverage at each of the sites, thereby determining an aggregate coverage-weighted normalized methylation difference score. Also provided herein are sensitive methods for using genomic DNA methylation levels to distinguish cancer cells from normal cells and to classify different cancer types according to their tissues of origin.

1 - 10 . (canceled) 
     
     
         11 . A method for distinguishing an aberrant methylation level for DNA from a first cell type, the method comprising
 (a) providing, for a plurality of CpG sites in baseline genomic DNA from two or more normal individual baseline organisms, a mean methylation level and a standard deviation of methylation level for each CpG site in the baseline genomic DNA;   (b) providing a test data set comprising:
 methylation states for the plurality of CpG sites from a first test genomic DNA from an individual test organism, wherein the CpG sites are derived from a sample, 
   (c) determining, for each of the CpG sites, the methylation difference between the first test genomic DNA and the baseline genomic DNA, thereby providing a normalized methylation difference for each CpG site; and   (d) converting the normalized methylation difference for each CpG site into a one-sided p-value;   (e) determining an aggregate methylation score for the combination of one-sided p-values for each CpG site for the first test genomic DNA.   
     
     
         12 . The method of  claim 11 , wherein (a) comprises providing methylation states for the plurality of CpG sites in the baseline genomic DNA from the two or more normal individual organisms, and determining, for each of the CpG sites, the mean methylation level and standard deviation of methylation level for the baseline genomic DNA. 
     
     
         13 . The method of  claim 11 , further comprising
 providing a second test data set comprising:
 methylation states for the plurality of CpG sites from a second test genomic DNA from the individual test organism, and wherein the CpG sites are derived from a sample; 
   determining, for each of the CpG sites, the methylation difference between the second test genomic DNA and the baseline genomic DNA, thereby providing a normalized methylation difference for each CpG site for the second test genomic DNA; and   converting the normalized methylation difference for each CpG site for the second test genomic DNA into a one-sided p-value;   determining an aggregate methylation score for the combination of one-sided p-values for each CpG site for the second test genomic DNA; and   comparing the aggregate methylation score of the first test genomic DNA and the second test genomic DNA to determine whether or not a change has occurred in the aggregate methylation score between the first and second test genomic DNA.   
     
     
         14 . The method of  claim 11 , further comprising:
 (f) providing a training data set comprising:
 methylation states for the plurality of CpG sites from training genomic DNA from two or more normal individual training organisms, wherein the CpG sites are derived from a plurality of different cell types from the normal individual training organisms; 
   (g) determining, for each of the CpG sites, the methylation difference between each training genomic DNA from the normal individual training organisms and the baseline genomic DNA, thereby providing a normalized methylation difference for each CpG site for each training genomic DNA;   (h) converting the normalized methylation difference for each CpG site into a one-sided p-value; and   (i) determining an aggregate methylation score for the combination of one-sided p-values for each CpG site for each training genomic DNA;   (j) using the aggregate methylation score for each training genomic DNA to calculate a mean aggregate methylation score and a standard deviation of the aggregate methylation scores for the training genomic DNA, to result in a distribution of the aggregate methylation scores for the training genomic DNA; and   (k) evaluating the aggregate methylation score of the first test genomic DNA against the distribution of the aggregate methylation scores for the training genomic DNA.   
     
     
         15 . The method of  claim 11 , wherein the normalized methylation difference at a particular CpG site is determined according to the formula: 
       
         
           
             
               
                 Z 
                 i 
               
               = 
               
                 
                   
                     χ 
                     i 
                   
                   - 
                   
                     μ 
                     i 
                   
                 
                 
                   σ 
                   i 
                 
               
             
           
         
         wherein Z i  represents a normalized methylation difference for a particular CpG site identified as i, χ i  represents the methylation level at CPG site i in the first test genomic DNA or the training genomic DNA, μ i  represents the mean methylation level at CpG site i in the baseline genome, and σ i  represents the standard deviation of methylation levels at CpG site i in the baseline genomic DNA. 
       
     
     
         16 . The method of  claim 14 , wherein the evaluating of step (k) determines a normalized methylation score difference according to the formula: 
       
         
           
             
               
                 Z 
                 
                   M 
                   ⁢ 
                   S 
                 
               
               = 
               
                 
                   
                     M 
                     ⁢ 
                     S 
                   
                   - 
                   
                     μ 
                     
                       M 
                       ⁢ 
                       S 
                     
                   
                 
                 
                   σ 
                   
                     M 
                     ⁢ 
                     S 
                   
                 
               
             
           
         
         wherein Z MS  represents a normalized methylation score difference, MS represents the aggregate methylation score of the first test genomic DNA, μ MS  represents the mean methylation score for the training set of normal genomic DNA, and σ MS  represents the standard deviation of aggregate methylation scores for the training set of normal genomic DNA. 
       
     
     
         17 . The method of  claim 11 , wherein the aggregate methylation score (MS) is determined according to the formula: 
       
         
           
             
               
                 M 
                 ⁢ 
                 S 
               
               = 
               
                 
                   - 
                   2 
                 
                 ⁢ 
                 
                   
                     ∑ 
                     
                       i 
                       = 
                       1 
                     
                     k 
                   
                   ⁢ 
                   
                     ln 
                     ⁡ 
                     
                       ( 
                       
                         p 
                         i 
                       
                       ) 
                     
                   
                 
               
             
           
         
         wherein p i  represents the one-sided p-value at site i, and k represents the total number of CpG sites. 
       
     
     
         18 . The method of  claim 11 , wherein the aggregate methylation score (MS) is determined according to the formula: 
       
         
           
             
               
                 M 
                 ⁢ 
                 S 
               
               = 
               
                 
                   - 
                   2 
                 
                 ⁢ 
                 
                   
                     ∑ 
                     
                       i 
                       = 
                       1 
                     
                     k 
                   
                   ⁢ 
                   
                     ln 
                     ⁡ 
                     
                       ( 
                       
                         
                           w 
                           i 
                         
                         ⁢ 
                         
                           p 
                           i 
                         
                       
                       ) 
                     
                   
                 
               
             
           
         
         wherein p i  represents the one-sided p-value at site i, k represents the total number of CpG sites, and w i  represents coverage of the site i. 
       
     
     
         19 . The method of  claim 11 , wherein the sample from the individual test organism comprises circulating tumor DNA and circulating non-tumor DNA. 
     
     
         20 . The method of  claim 11 , wherein the sample comprises cell-free DNA from blood. 
     
     
         21 . The method of  claim 11 , wherein the individual test organism is a pregnant female and the first test genomic DNA comprises genomic DNA derived from somatic cells of the female and genomic DNA derived from somatic cells of prenatal offspring of the female. 
     
     
         22 . The method of  claim 11 , wherein the providing of the sample in step (b) comprises targeted selection of a subset of genomic DNA fragments comprising a set of predetermined target CpG sites. 
     
     
         23 . The method of  claim 11 , wherein the providing of the sample in step (b) further comprises treating the subset of genomic DNA fragments with bisulfite. 
     
     
         24 . The method of  claim 11 , wherein the providing of the sample in step (b) comprises detecting methylation states for the CpG sites. 
     
     
         25 . The method of  claim 11 , wherein the detecting methylation states for the CpG sites comprises a sequencing technique that sequentially identifies nucleotides in the first test genomic DNA. 
     
     
         26 . The method of  claim 16 , wherein the first test genomic DNA is classified as having an aberrant methylation level if the value of Z MS  is greater than 3. 
     
     
         27 - 35 . (canceled)