METHODS, SYSTEMS, AND COMPUTER READABLE MEDIA FOR MAKING BASE CALLS IN NUCLEIC ACID SEQUENCING
A method for nucleic acid sequencing includes: receiving a signal comprising measurements of a parameter measured in response to a plurality of nucleotide flows flowed in a space comprising a sample nucleic acid; normalizing the signal to obtain a normalized signal; adaptively normalizing the normalized signal to obtain an adaptively normalized signal; and predicting a sequence of base calls corresponding to the sample nucleic acid using the adaptively normalized signal.
1 . A method for nucleic acid sequencing, comprising:
receiving a signal comprising measurements of a parameter measured in response to a plurality of nucleotide flows flowed in a space comprising a sample nucleic acid;
normalizing the signal to obtain a normalized signal;
adaptively normalizing the normalized signal to obtain an adaptively normalized signal; and
predicting a sequence of base calls corresponding to the sample nucleic acid using the adaptively normalized signal.
2 . The method of claim 1 , wherein the parameter is a physical parameter that is substantially proportional to a number of nucleotide incorporations having resulted in the space comprising the sample nucleic acid in response to each of the nucleotide flows.
3 . The method of claim 1 , wherein the parameter is derived from one or more voltage measurements indicative of an hydrogen ion concentration in the space comprising the sample nucleic acid.
4 . The method of claim 1 , wherein normalizing the signal comprises scaling the measurements using a reference value.
5 . The method of claim 4 , wherein the reference value comprises an average of a plurality of measurements of the parameter measured in response to one or more initial nucleotide flows flowed in a space comprising a reference nucleic acid designed such that a pre-determined number of nucleotide incorporations would be expected to result from each of the one or more initial nucleotide flows.
6 . The method of claim 1 , wherein adaptively normalizing the normalized signal comprises applying a time-varying additive correction to the normalized signal.
7 . The method of claim 6 , wherein the time-varying additive correction is obtained by (1) fitting an additive term for a measurement corresponding to nucleotide flow i to a difference between that measurement and a corresponding predicted value for nucleotide flow i whenever the corresponding predicted value for nucleotide flow i does not exceed a minimal threshold for additive correction, (2) calculating a median from the additive terms for each of a plurality of consecutive blocks of nucleotide flows, and (3) linearly interpolating between the centers of each of the consecutive blocks of nucleotide flows to obtain the time-varying additive correction.
8 . The method of claim 1 , wherein adaptively normalizing the normalized signal comprises applying a time-varying multiplicative correction to the normalized signal.
9 . The method of claim 6 , wherein adaptively normalizing the normalized signal further comprises applying a time-varying multiplicative correction to the normalized signal to which the time-varying additive correction has been applied.
10 . The method of claim 9 , wherein the time-varying multiplicative correction is obtained by (1) fitting a multiplicative term for a measurement corresponding to nucleotide flow i to a ratio between (i) a difference between that measurement and the additive correction for nucleotide flow i and (ii) the corresponding predicted value for nucleotide flow i whenever the corresponding predicted value for nucleotide flow i exceeds a minimal threshold for multiplicative correction, (2) calculating a median from the multiplicative terms for each of a plurality of consecutive blocks of nucleotide flows, and (3) linearly interpolating between the centers of each of the consecutive blocks of nucleotide flows to obtain the time-varying multiplicative correction.
11 . The method of claim 1 , further comprising repeating the adaptively normalizing and predicting steps in successive iterations such that predicted measurement values generated during the predicting step are used in subsequent iterations of the adaptively normalizing step.
12 . The method of claim 11 , wherein the successive iterations steps are progressively longer such that a first iteration of the adaptively normalizing and predicting steps is performed for a first set of nucleotide flows, and a second iteration of the adaptively normalizing and predicting steps is performed for a second set of nucleotide flows that is larger than and includes the first set of nucleotide flows.
13 . The method of claim 12 , wherein a third iteration of the adaptively normalizing and predicting steps is performed for a third set of nucleotide flows that is larger than and includes the second set of nucleotide flows.
14 . The method of claim 1 , wherein predicting a sequence of base calls is performed without separately modeling droop.
15 . The method of claim 11 , wherein the successive iterations are performed such that results from a subset of earlier flows are reused in successive iterations without additional processing whereas results for a subset of later flows are determined in subsequent iterations of the adaptively normalizing step.
16 . The method of claim 11 , wherein the successive iterations of the adaptively normalizing and predicting steps are preceded by a single initial basecalling step that provides an initial sequence of base calls for an initial set of nucleotide flows.
17 . The method of claim 11 , further comprising reusing, during the successive iterations of the adaptively normalizing and predicting steps, results of the predicting step obtained in a prior iteration for a subset of nucleotide flows while performing the predicting steps for a subset of additional nucleotide flows in the current iteration.
18 . A non-transitory machine-readable storage medium comprising instructions which, when executed by a processor, cause the processor to perform a method for nucleic acid sequencing comprising:
receiving a signal comprising measurements of a parameter measured in response to a plurality of nucleotide flows flowed in a space comprising a sample nucleic acid;
normalizing the signal to obtain a normalized signal;
adaptively normalizing the normalized signal to obtain an adaptively normalized signal; and
predicting a sequence of base calls corresponding to the sample nucleic acid using the adaptively normalized signal.
19 . The non-transitory machine-readable storage medium of claim 18 , wherein adaptively normalizing the normalized signal comprises applying a time-varying additive correction to the normalized signal and applying a time-varying multiplicative correction to the normalized signal.
20 . A system, comprising:
a machine-readable memory; and
a processor configured to execute machine-readable instructions, which, when executed by the processor, cause the system to perform steps including:
receiving a signal comprising measurements of a parameter measured in response to a plurality of nucleotide flows flowed in a space comprising a sample nucleic acid;
normalizing the signal to obtain a normalized signal;
adaptively normalizing the normalized signal to obtain an adaptively normalized signal; and
predicting a sequence of base calls corresponding to the sample nucleic acid using the adaptively normalized signal.