Molecular testing of multiple pregnancies

Abstract

Methods, systems, and apparatus are provided for determining zygosity of a multiple-fetus pregnancy using a biological sample taken from the mother. The fetal and maternal DNA in the sample (e.g. plasma) can be analyzed for a particular chromosomal region to identify genetic differences in the fetuses. For example, a normalized parameter for the measure of a primary or secondary allele can show variances for different chromosomal regions when fetuses are dizygotic. Such a variance can be determined relative to an expected value if the fetuses were genetically identical. Statistical methods are provided for analyzing the variation of the normalized parameters to determine fetal DNA concentration and the maternal-fetal mixed genotype at various loci. Parental genotype and haplotype information can also be used to identify inheritance of different parental haplotypes to indicate genetic differences among the fetuses.

What is claimed is: 
     
         1 . A method for analyzing a biological sample of a female pregnant with a plurality of fetuses to determine whether at least two fetuses of a pregnant female are dizygotic, the biological sample comprising fetal and maternal DNA, the method comprising:
 receiving sequenced reads obtained by performing massively parallel sequencing of DNA fragments in the biological sample;   aligning, by a computer system, the sequenced reads to a reference human genome to identify locations and alleles of the DNA fragments, thereby obtaining data about alleles of the DNA fragments;   determining, by the computer system, a genotype of the pregnant female at each of one or more first loci within a first chromosomal region, the pregnant female being homozygous at each of the one or more first loci or being heterozygous at each of the one or more first loci, wherein each of the first loci exhibits a respective primary allele and a respective secondary allele in the biological sample, wherein the respective primary allele is more abundant than the respective secondary allele for each of the first loci;   measuring, using the data at the one or more first loci, a first amount of the one or more primary alleles and/or a second amount of the one or more secondary alleles in the biological sample;   obtaining a normalized parameter for the first amount or the second amount;   comparing the normalized parameter to a cutoff value to determine if the normalized parameter is statistically different from an expected value if the fetuses are genetically identical for the first chromosomal region, the expected value being obtained from a measurement of the biological sample; and   determining that the at least two fetuses of the pregnant female are dizygotic based on the normalized parameter exceeding the cutoff value.   
     
     
         2 . The method of  claim 1 , further comprising:
 at each of the first loci, detecting the respective primary allele and the respective secondary allele in the biological sample.   
     
     
         3 . The method of  claim 1 , wherein obtaining the normalized parameter includes:
 measuring a third amount of one or more sequences from one or more loci within a different chromosomal region; and   using the third amount to calculate the normalized parameter.   
     
     
         4 . The method of  claim 1 , wherein obtaining the normalized parameter includes:
 measuring, at the one or more first loci, the first amount of the one or more primary alleles and the second amount of the one or more secondary alleles in the biological sample;   determining a first parameter from the first amount and the second amount, the first parameter providing a relative amount between the first amount and the second amount; and   using the first parameter as the normalized value of the first amount.   
     
     
         5 . The method of  claim 1 , wherein obtaining the normalized parameter includes:
 performing the measuring step using a calibrated process; and   using the first amount or the second amount as the normalized parameter.   
     
     
         6 . The method of  claim 1 , wherein the expected value is a fetal DNA concentration in the biological sample. 
     
     
         7 . The method of  claim 6 , further comprising:
 measuring the fetal DNA concentration using one or more epigenetic markers.   
     
     
         8 . The method of  claim 7 , wherein the one or more epigenetic markers include one or more DNA methylation markers. 
     
     
         9 . The method of  claim 6 , further comprising:
 calculating the fetal DNA concentration using genetic markers by:   measuring a third amount of DNA fragments having a fetal-specific sequence selected from one or more fetal-specific sequences, wherein all of the fetuses have the fetal-specific sequence;   obtaining a normalized value for the third amount; and   using the normalized value as the fetal DNA concentration.   
     
     
         10 . The method of  claim 9 , wherein the fetuses are all males, and wherein the one or more fetal-specific sequences are on the Y chromosome. 
     
     
         11 . The method of  claim 9 , wherein the fetal-specific sequence is the RHD gene, wherein the mother is RhD-negative, and wherein all of the fetuses are RhD-positive. 
     
     
         12 . The method of  claim 9 , further comprising:
 identifying one or more second loci at which the fetuses have a respective first allele and the mother does not have the respective first allele, wherein the fetal-specific sequences are the respective first alleles, wherein obtaining the normalized value for the third amount includes:   measuring a total amount of alleles at the one or more second loci;   calculating the fetal DNA concentration from a ratio of the third amount and the total amount.   
     
     
         13 . The method of  claim 1 , wherein the one or more loci includes the locus of the RHD gene, the mother being homozygous for an allele represented by the absence of the RHD gene, and wherein at least one of the fetuses is RhD-positive. 
     
     
         14 . The method of  claim 1 , further comprising:
 determining a genotype of the pregnant female at each of one or more second loci within a second chromosomal region, the pregnant female being homozygous at each of the one or more second loci or being heterozygous at each of the one or more second loci, wherein each of the second loci exhibit a respective primary allele and a respective secondary allele in the biological sample;   measuring, at the one or more second loci, a third amount of the one or more primary alleles and/or a fourth amount of the one or more secondary alleles in the biological sample;   obtaining a second normalized parameter for the third amount or the fourth amount; and   using the second normalized parameter to obtain the expected value.   
     
     
         15 . The method of  claim 14 , wherein the second normalized parameter is used as the expected value. 
     
     
         16 . The method of  claim 14 , further comprising:
 calculating additional normalized parameters for other chromosomal regions;   computing a first statistical value from the normalized parameters of a first group of chromosomal regions; and   using the first statistical value as the expected value.   
     
     
         17 . The method of  claim 16 , wherein the respective primary allele is more abundant than the respective secondary allele for all of the loci of the chromosomal regions of the first group. 
     
     
         18 . The method of  claim 16 , wherein the first statistical value is a location of a peak in the values of the normalized parameters in a histogram of the normalized parameters for the first group of chromosomal regions. 
     
     
         19 . The method of  claim 16 , further comprising:
 computing a second statistical value from the parameters of a second group of chromosomal regions that includes the first chromosomal region,   wherein comparing the normalized parameter to the cutoff value includes:
 comparing the second statistical value to the cutoff value. 
   
     
     
         20 . The method of  claim 19 , further comprising:
 identifying the first and second groups of chromosomal regions by analyzing a histogram of the calculated parameters, wherein the first group of chromosomal regions corresponds to a first peak in the histogram and the second group of chromosomal regions corresponds to a second peak in the histogram.   
     
     
         21 . The method of  claim 14 , further comprising:
 calculating additional normalized parameters for other chromosomal regions,   wherein comparing the normalized parameter to the cutoff value to determine if the normalized parameter is statistically different from the expected value is accomplished by:
 computing a variance in the values of the normalized parameters; and 
 comparing the variance to a threshold value. 
   
     
     
         22 . The method of  claim 21 , wherein the variation is the standard deviation in the normalized parameters. 
     
     
         23 . The method of  claim 1 , further comprising performing massively parallel sequencing of the DNA fragments in the biological sample. 
     
     
         24 . A computer product comprising a non-transitory computer readable medium storing a plurality of instructions that when executed control a computer system to analyze a biological sample of a female pregnant with a plurality of fetuses to determine whether at least two fetuses of a pregnant female are dizygotic, the biological sample comprising fetal and maternal DNA, the instructions comprising:
 receiving sequenced reads obtained by performing massively parallel sequencing of DNA fragments in the biological sample;   aligning the sequenced reads to a reference human genome to identify locations and alleles of the DNA fragments, thereby obtaining data about alleles of the DNA fragments;   determining a genotype of the pregnant female at each of one or more first loci within a first chromosomal region, the pregnant female being homozygous at each of the one or more first loci or being heterozygous at each of the one or more first loci, wherein each of the first loci exhibits a respective primary allele and a respective secondary allele in the biological sample, wherein the respective primary allele is more abundant than the respective secondary allele for each of the first loci;   determining, using the data at the one or more first loci, a first amount of the one or more primary alleles and/or a second amount of the one or more secondary alleles in the biological sample;   obtaining a normalized parameter for the first amount or the second amount; and   comparing the normalized parameter to a cutoff value to determine if the normalized parameter is statistically different from an expected value if the fetuses are genetically identical for the first chromosomal region, the expected value being obtained from a measurement of the biological sample; and   determining that at least two fetuses of the pregnant female are dizygotic based on the normalized parameter exceeding the cutoff value.   
     
     
         25 . A method for analyzing a biological sample of a female pregnant with a plurality of fetuses to determine whether at least two fetuses of a pregnant female are dizygotic, the biological sample comprising fetal and maternal DNA, the method comprising:
 receiving sequenced reads obtained by performing massively parallel sequencing of DNA fragments in the biological sample;   aligning, by a computer system, the sequenced reads to a reference human genome to identify locations and alleles of the DNA fragments, thereby obtaining data about alleles of the DNA fragments;   measuring, using the data at one or more first loci, a first amount of one or more fetal-specific sequences in the biological sample;   obtaining a normalized parameter for the first amount;   comparing the normalized parameter to a cutoff value to determine if the normalized parameter is statistically different from an expected value if the fetuses are genetically identical for a first chromosomal region, the expected value being obtained from a measurement of the biological sample; and   determining that the at least two fetuses of the pregnant female are dizygotic based on the normalized parameter exceeding the cutoff value.   
     
     
         26 . The method of  claim 25 , wherein a fetal-specific sequence of the one or more fetal-specific sequences is on the Y chromosome. 
     
     
         27 . The method of  claim 25 , wherein a fetal-specific sequence of the one or more fetal-specific sequences is the RHD gene, and wherein the pregnant female is RhD-negative. 
     
     
         28 . The method of  claim 25 , wherein the expected value is a fetal DNA concentration in the biological sample. 
     
     
         29 . The method of  claim 25 , wherein the one or more first loci are within the first chromosomal region, the method further comprising:
 measuring, at one or more second loci, a second amount of one or more additional fetal-specific sequences in the biological sample, the second loci being within a second chromosomal region different than the first chromosomal region;   obtaining an additional normalized parameter for the second amount; and   using the additional normalized parameter to obtain the expected value.   
     
     
         30 . A method for analyzing a biological sample of a female pregnant with a plurality of fetuses to determine whether at least two fetuses of the pregnant female are dizygotic, the biological sample comprising fetal and maternal DNA, the method comprising:
 receiving sequenced reads obtained by performing massively parallel sequencing of DNA fragments in the biological sample;   aligning, by a computer system, the sequenced reads to a reference human genome to identify locations and alleles of the DNA fragments, thereby obtaining data about alleles of the DNA fragments;   for each of a plurality of chromosomal regions:
 at each of one or more loci in the respective chromosomal region:
 measuring, using the data, one or more alleles in the biological sample; and 
 determining, using the data, a respective amount of each measured allele; 
 
 based on the respective amounts of the measured alleles, determining whether at least two of the fetuses have inherited a different haplotype of the respective chromosomal region from a first parent; 
   determining a first amount of the chromosomal regions where at least two of the fetuses have inherited a different haplotype from the first parent; and   comparing the first amount to a cutoff value to determine whether the at least two of the fetuses are dizygotic based on the first amount exceeding the cutoff value.   
     
     
         31 . The method of  claim 30 , wherein the first amount is a proportion. 
     
     
         32 . The method of  claim 30 , the method further comprising:
 determining the two haplotypes of the first parent at a plurality of loci for a first chromosomal region, wherein determining that at least two of the fetuses have inherited a different haplotype of the first chromosomal region from the first parent includes:
 identifying a first locus and a second locus in the first chromosomal region at which the first parent is heterozygous; 
 detecting in the biological sample a first haplotype of the first parent at the first locus; and 
 detecting in the biological sample a second haplotype of the first parent at the second locus. 
   
     
     
         33 . The method of  claim 32 , wherein the first parent is the mother, wherein the father is homozygous at the first locus for a first allele, and the mother is heterozygous for the first allele and a second allele at the first locus, the first allele being on the first haplotype and the second allele being on the second haplotype of the mother, wherein detecting in the biological sample the first haplotype of the first parent at the first locus includes:
 determining that the respective amount of the first allele measured at the first locus is greater than the respective amount of the second allele measured at the second locus by a statistically significant amount;   wherein the father is homozygous at the second locus for a third allele, and the mother is heterozygous for the third allele and a fourth allele at the second locus, the third allele being on the first haplotype and the fourth allele being on the second haplotype of the mother, wherein detecting in the biological sample the second haplotype of the first parent at the second locus includes:
 determining that the respective amount of the fourth allele measured at the second locus is greater than the respective amount of the third allele measured at the second locus by a statistically significant amount. 
   
     
     
         34 . The method of  claim 32 , wherein the first parent is the father, wherein the second parent is homozygous at the first locus for a first allele, and the first parent is heterozygous for the first allele and a second allele at the first locus, wherein detecting in the biological sample the first haplotype of the first parent at the first locus includes:
 detecting in the biological sample the second allele at the first locus.   
     
     
         35 . The method of  claim 32 , wherein the first locus and the second locus are close enough that the probability of recombination between the first locus and the second locus is less than 0.1%. 
     
     
         36 . The method of  claim 30 , wherein the first parent is the father, the method further comprising:
 identifying a first locus in a first chromosomal region where the first parent is heterozygous for a first allele and a second allele, both of which are not present in the second parent at the first locus,   wherein determining that two of the fetuses have inherited a different haplotype of the first chromosomal region from the first parent includes:
 detecting the first allele and the second allele at the first locus. 
   
     
     
         37 . The method of  claim 36 , wherein the first allele and the second allele contain different numbers of a short tandem repeat. 
     
     
         38 . A method of determining a fetal DNA percentage in a biological sample from a pregnant female with at least two fetuses, the biological sample comprising fetal and maternal DNA, the method comprising:
 receiving sequenced reads obtained by performing massively parallel sequencing of DNA fragments in the biological sample;   aligning, by a computer system, the sequenced reads to a reference human genome to identify locations and alleles of the DNA fragments, thereby obtaining data about alleles of the DNA fragments;   creating a histogram by:
 for each of a plurality of chromosomal regions:
 identifying, using the data, one or more loci in the respective chromosomal region at which a respective first allele and a respective second allele are detected in the biological sample; 
 measuring, using the data at the one or more loci, a first amount of the one or more first alleles and/or a second amount of the one or more second alleles in the biological sample; and 
 obtaining a normalized parameter for the first amount or the second amount; and 
 
 incrementing counters based on a number of chromosomal regions with specified values for the normalized parameter; 
   fitting a linear combination of probability distributions to the histogram, where the fetal DNA percentage is an input to the linear combination of probability distributions; and   varying an input fetal DNA percentage to find an optimal fetal DNA percentage that optimizes a fit of the linear combination of probability distributions to the histogram.   
     
     
         39 . The method of  claim 38 , wherein the linear combination of probability distributions has two probability distributions, where both probability distributions are dependent on the fetal DNA percentage. 
     
     
         40 . The method of  claim 38 , wherein the linear combination of probability distributions has four probability distributions, where two of the probability distributions are dependent on the fetal DNA percentage. 
     
     
         41 . The method of  claim 38 , wherein the normalized parameter is computed from the first amount and the second amount, the normalized parameter providing a relative amount between the first amount and the second amount. 
     
     
         42 . The method of  claim 38 , wherein the probability distributions are binomial distributions or Gaussian distributions. 
     
     
         43 . The method of  claim 38 , further comprising:
 based on the optimal fetal DNA percentage, identifying the probability distribution corresponding to loci at which the mother is homozygous and at least one of the fetuses is heterozygous;   fitting a multi-component mixture model to the identified probability distribution, the multi-component mixture model including a mixture coefficient for each of a plurality of components; and   determining that at least two of the fetuses are dizygotic when at least two of the mixture coefficients are above a threshold.   
     
     
         44 . The method of  claim 43 , wherein the threshold is dependent on at least one of the mixture coefficients. 
     
     
         45 . The method of  claim 43 , wherein locations of the peaks of the components of the mixture model are constrained to have a separation gap that exceeds a predetermined value. 
     
     
         46 . A method for analyzing a biological sample of a pregnant female to determine whether at least two fetuses of the pregnant female are dizygotic, the biological sample comprising fetal and maternal DNA, the method comprising:
 receiving sequenced reads obtained by performing massively parallel sequencing of DNA fragments in the biological sample;   aligning, by a computer system, the sequenced reads to a reference human genome to identify locations and alleles of the DNA fragments, thereby obtaining data about alleles of the DNA fragments;   creating a histogram by:
 for each of a plurality of chromosomal regions:
 identifying, using the data, one or more loci in the respective chromosomal region at which a respective first allele and a respective second allele are detected in the biological sample; 
 measuring, using the data at the one or more loci, a first amount of the one or more first alleles and/or a second amount of the one or more second alleles in the biological sample; and 
 obtaining a normalized parameter for the first amount or the second amount; and 
 
 incrementing counters based on a number of chromosomal regions with specified values for the normalized parameter; 
   identifying chromosomal regions corresponding to loci at which the mother is homozygous and at least one of the fetuses is heterozygous or corresponding to loci at which the mother is heterozygous and at least one of the fetuses is homozygous;   fitting a multi-component mixture model to the histogram corresponding to the identified chromosomal regions, the multi-component mixture model including a mixture coefficient for each of a plurality of components; and   determining whether at least two of the fetuses are dizygotic using at least two of the mixture coefficients.   
     
     
         47 . The method of  claim 46 , wherein the multi-component mixture model has three components. 
     
     
         48 . The method of  claim 46 , wherein the multi-component mixture model is a Gaussian mixture model. 
     
     
         49 . The method of  claim 46 , further comprising:
 determining genotypes of the fetuses for the one or more loci of a first chromosomal region by identifying the component of the mixture model with a highest overlap with the corresponding normalized parameter for the first chromosomal region.   
     
     
         50 . The method of  claim 46 , wherein at least two of the fetuses are determined to be dizygotic when at least two of the mixture coefficients are above a threshold. 
     
     
         51 . The method of  claim 46 , wherein locations of the peaks of the components of the mixture model are constrained to have a separation gap that exceeds a predetermined value. 
     
     
         52 . The method of  claim 46 , wherein identifying the chromosomal regions includes:
 fitting a linear combination of probability distributions to the histogram for a given fetal DNA percentage of the biological sample; and   identifying a probability distribution corresponding to loci at which the mother is homozygous and at least one of the fetuses is heterozygous or corresponding to loci at which the mother is heterozygous and at least one of the fetuses is homozygous, wherein the multi-component mixture model is fit to the identified probability distribution.   
     
     
         53 . The method of  claim 46 , further comprising:
 determining the given fetal DNA percentage using one or more epigenetic markers.   
     
     
         54 . The method of  claim 53 , wherein the one or more epigenetic markers include one or more DNA methylation markers.