US2022336049A1PendingUtilityA1
Systems and methods for determining effects of therapies and genetic variation on polyadenylation site selection
Est. expiryAug 8, 2038(~12.1 yrs left)· nominal 20-yr term from priority
G16B 20/00C12Q 2600/106G16B 30/00G06N 20/00C12N 2310/11G16B 20/20C12N 15/111C12N 2320/11C12Q 1/6869C12N 2830/50G16B 40/20C12Q 1/6883G06N 3/008C12N 15/113C12Q 2600/156G16B 25/00G16B 40/00
78
PatentIndex Score
0
Cited by
0
References
0
Claims
Abstract
The present disclosure provides systems and methods for determining effects of genetic variants on selection of polyadenylation sites (PAS) during polyadenylation processes. In an aspect, the present disclosure provides a polyadenylation code, a computational model that can predict alternative polyadenylation patterns from transcript sequences. A score can be calculated that describes or corresponds to the strength of a PAS, or the efficiency in which it is recognized by the 3′-end processing machinery. The polyadenylation model may be used, for example, to assess the effects of anti-sense oligonucleotides to alter transcript abundance. As another example, the polyadenylation model may be used to scan the 3′-UTR of a human genome to find potential PAS.
Claims
exact text as granted — not AI-modified1 .- 94 . (canceled)
95 . A computer-implemented method for determining an effect of an antisense oligonucleotide, the method comprising:
(a) providing a plurality of genomic sequences, wherein the plurality of genomic sequences comprises (1) a reference sequence and (2) a variant sequence obtained by computer processing the reference sequence based at least in part on the antisense oligonucleotide, wherein the reference sequence is (i) derived from a human genome, (ii) obtained by sequencing deoxyribonucleic acid (DNA) or ribonucleic acid (RNA) of a bodily sample obtained or derived from a subject, or (iii) a genetic aberration thereof, and wherein the antisense oligonucleotide is complementary to at least a portion of the reference sequence; (b) for each of the plurality of genomic sequences:
i. identifying a plurality of candidate polyadenylation sites in the genomic sequence;
ii. extracting a polyadenylation feature vector for each of the plurality of candidate polyadenylation sites, wherein each of the polyadenylation feature vectors comprises a set of features determined based at least in part on a set of nucleotides in the genomic sequence; and
iii. applying a trained machine learning algorithm comprising a neural network to the plurality of polyadenylation feature vectors to determine a set of preferences p 1 , p 2 , . . . , p n for the plurality of candidate polyadenylation sites, wherein each preference p, among the set of preferences p 1 , p 2 , . . . , p n indicates a probability of selection of an ith candidate polyadenylation site among the plurality of candidate polyadenylation sites; and
(c) computer processing the plurality of sets of preferences for each of the plurality of genomic sequences with each other to determine the effect of the antisense oligonucleotide on the plurality of candidate polyadenylation sites, wherein the antisense oligonucleotide modulates polyadenylation of at least one of the plurality of candidate polyadenylation sites.
96 . The method of claim 95 , wherein determining the set of preferences for each of the plurality of genomic sequences comprises:
for each of the plurality of candidate polyadenylation sites, applying the trained machine learning algorithm to the plurality of polyadenylation feature vectors to determine an intermediate representation r i for an ith candidate polyadenylation site, the intermediate representation comprising a numerical value; and computer processing the set of intermediate representations r 1 , r 2 , . . . , r n for the plurality of candidate polyadenylation sites to determine the set of preferences p 1 , p 2 , . . . , p n corresponding to the plurality of candidate polyadenylation sites.
97 . The method of claim 96 , wherein the neural network comprises a deep neural network, a convolutional neural network, a recurrent neural network, or a long short-term memory recurrent neural network.
98 . The method of claim 97 , wherein the neural network comprises the convolutional neural network.
99 . The method of claim 96 , wherein the intermediate representation for the ith candidate polyadenylation site comprises a numerical value and wherein computer processing the set of intermediate representations comprises applying a softmax function to the set of intermediate representations r 1 , r 2 , . . . , r n for the plurality of candidate polyadenylation sites to determine the set of preferences p 1 , p 2 , . . . , p n for the plurality of candidate polyadenylation sites.
100 . The method of claim 96 , wherein the intermediate representation for the ith candidate polyadenylation site comprises a numerical value r i , and wherein computer processing the set of intermediate representations comprises determining each preference p i of the set of preferences as
p
i
←
exp
(
r
i
)
exp
(
?
)
+
exp
(
?
)
+
…
+
exp
(
?
)
,
?
indicates text missing or illegible when filed
wherein exp is an exponential function or a numerical approximation of an exponential function.
101 . The method of claim 96 , wherein the intermediate representation for the ith candidate polyadenylation site comprises a numerical value r i , and wherein computer processing the set of intermediate representations comprises determining each preference p i of the set of preferences as
p
i
←
relu
(
r
i
)
relu
(
?
)
+
relu
(
?
)
+
…
+
relu
(
?
)
,
?
indicates text missing or illegible when filed
wherein relu is a rectified linear function.
102 . The method of claim 96 , wherein the intermediate representation for the ith candidate polyadenylation site comprises a numerical value r i , and wherein computer processing the set of intermediate representations comprises determining each preference p i of the set of preferences as
p
i
←
m
(
r
i
)
m
(
?
)
+
m
(
?
)
+
…
+
m
(
?
)
,
?
indicates text missing or illegible when filed
wherein m( ) is a non-negative monotonic function.
103 . The method of claim 95 , wherein the genetic aberration comprises a single nucleotide variant (SNV) or an insertion or deletion (indel).
104 . The method of claim 95 , wherein at least one of the set of nucleotides is located within about 100 nucleotides of a location in the genomic sequence of the candidate polyadenylation site.
105 . The method of claim 95 , wherein each of the plurality of polyadenylation feature vectors comprises one or more of:
(a) a subsequence of the genomic sequence encoded using a 1-of-4 binary vector for a nucleotide selected from adenine (A), thymine (T), cytosine (C), and guanine (G); (b) a subsequence of the genomic sequence encoded using a 1-of-4 binary vector for a nucleotide selected from adenine (A), uracil (U), cytosine (C), and guanine (G); (c) a set of binary components; (d) a set of categorical components; (e) a set of integer components; and (f) a set of real-valued components.
106 . The method of claim 105 , wherein at least one of the set of binary components comprises a value indicative of the presence of a cleavage factor sequence in the candidate polyadenylation site, or a value indicative of the absence of a cleavage factor sequence in the candidate polyadenylation site.
107 . The method of claim 105 , wherein at least one of the set of real-valued components comprises a log distance, in number of nucleotides in the genomic sequence, from (1) the candidate polyadenylation site to (2) a nearest different candidate polyadenylation site among the plurality of candidate polyadenylation sites.
108 . The method of claim 105 , wherein at least one of the set of binary components comprises a value indicative of the presence of a cleavage factor sequence adjacent to the candidate polyadenylation site or a value indicative of the absence of a cleavage factor sequence adjacent to the candidate polyadenylation site.
109 . The method of claim 105 , wherein the at least one of the plurality of polyadenylation feature vectors comprises a feature selected from the group consisting of:
a PolyA signal AAUAAA in a 5′-3′ region, a log distance between the candidate polyadenylation site and a nearest different candidate polyadenylation site among the plurality of candidate polyadenylation sites, a PolyA signal AUUAAA in a 5′-3′ region, a 1-mer C in a 5′-3′ region, a 1-mer U in a 5′-3′ region, a 2-mer AG in a 5′-3′ region, a 2-mer CA in a 3′-5′ region, a 3-mer AAA in a 3′-5′ region, a 3-mer UGU in a 5′-3′ region, a 3-mer UGU in a 5′-5′ region, a 4-mer AAAA in a 3′-5′ region, a cleavage factor Im UGUA in a 5′-5′ region, a PolyA signal CAAUAA in a 5′-3′ region, a PolyA signal AUAAAG in a 5′-3′ region, and a PolyA signal AGUAAA in a 5′-5′ region.
110 . The method of claim 95 , further comprising, for at least one of the plurality of genomic sequences, identifying a maximally preferred candidate polyadenylation site among the plurality of candidate polyadenylation sites, wherein the maximally preferred candidate polyadenylation site has a largest numerical value p max among the set of preferences p 1 , p 2 , . . . , p n .
111 . The method of claim 95 , wherein determining the set of preferences comprises:
providing a set of numerical parameters; and determining a multiplication product comprising at least one feature from at least one of the plurality of polyadenylation feature vectors and at least one numerical parameter of the set of numerical parameters.
112 . The method of claim 111 , wherein applying the trained machine learning algorithm comprises adjusting a numerical parameter of the set of numerical parameters to decrease a loss function.
113 . The method of claim 112 , wherein adjusting the numerical parameter of the set of numerical parameters comprises performing a gradient-based learning procedure.
114 . The method of claim 113 , wherein the gradient-based learning procedure comprises stochastic gradient descent.
115 . The method of claim 112 , wherein the loss function comprises a cross entropy function.
116 . The method of claim 95 , wherein a one-to-one correspondence exists between one or more of the plurality of candidate polyadenylation sites of the reference sequence and one or more of the plurality of candidate polyadenylation sites of the variant sequence, and wherein processing the plurality of sets of preferences comprises comparing each of at least one preference in the set of preferences of the reference sequence to the corresponding preference in the set of preferences of the variant sequence which is in one-to-one correspondence.
117 . The method of claim 116 , wherein (c) further comprises calculating a set of changes in preference Δp 1 , Δp 2 . . . , Δp n corresponding to the plurality of candidate polyadenylation sites of the reference sequence and the plurality of candidate polyadenylation sites of the variant sequence to determine the effect of the antisense oligonucleotide.
118 . The method of claim 95 , wherein the variant sequence is obtained by replacing one or more nucleotides of the at least the portion of the reference sequence with an N base, a uniform weighting of the 4 bases, or randomly selected bases.
119 . The method of claim 95 , wherein the plurality of polyadenylation feature vectors is indicative of a relative positioning of the plurality of candidate polyadenylation sites to determine the set of preferences.
120 . The method of claim 95 , wherein the determined effect of the antisense oligonucleotide comprises a decreased preference for one or more of the plurality of candidate polyadenylation sites.
121 . The method of claim 95 , wherein the determined effect of the antisense oligonucleotide comprises an increased preference for one or more of the plurality of candidate polyadenylation sites.
122 . The method of claim 95 , wherein the antisense oligonucleotide has a length of about 10 to about 50 nucleotides.
123 . The method of claim 95 , further comprising determining a tissue-specific effect of the antisense oligonucleotide based at least in part on whether a plurality of polyadenylation feature vectors of the plurality of candidate polyadenylation sites comprises tissue-specific polyadenylation features.
124 . The method of claim 123 , wherein the tissue-specific polyadenylation features comprise a feature selected from the group consisting of:
a 4-mer UUGU in a 5′-5′ region, a 3-mer UUG in a 3′-3′ region, a 4-mer CCCC in a 3′-3′ region, a 3-mer, UGU in a 5′-5′ region, a 4-mer, UCCC in a 3′-3′ region, a 4-mer, CGGC in a 5′-3′ region, a cis-element UUUGUA in a 5′-5′ region, a cleavage factor Im UGUA in a 5′-5′ region, a 3-mer, UUG in a 5′-5′ region, a 3-mer, AUC in a 5′-5′ region, a 3-mer, UCC in a 3′-3′ region, a 2-mer, UC in a 5′-5′ region, a 4-mer, AUCC in a 5′-5′ region, a 2-mer, UU in a 5′-5′ region, and a 3-mer CCU in a 3′-3′ region.Cited by (0)
No later patents cite this yet.
References (0)
No backward citations on record.