CN106446600A - CRISPR/Cas9-based sgRNA design method - Google Patents

Abstract

The invention relates to a CRISPR/Cas9-based sgRNA design method. The method is characterized by comprising the following steps of obtaining the value of enzyme digestion efficiency of sgRNA and corresponding Cas9; building a personalized sgRNA design model; measuring the quality of the built personalized sgRNA design model by applying an NDCG algorithm and updating a database; and designing sgRNA and giving out an assessment value of each sgRNA. Compared with the prior art, the method has the characteristics of high accuracy, complete characteristics, wide application range and wide data analysis range.

Description

A kind of method for designing of the sgRNA based on CRISPR/Cas9

Technical field

The present invention relates to gene editing research field, especially a kind of based on CRISPR/Cas9 gene editing technology The method for designing of sgRNA.

Background technology

With the development of molecular biology, people have had more further understanding for the constitution element of life, but raw The curative mechanism of the mechanism of life process, especially some disease there is also a lot of not understanding.Relation between gene and phenotype, gene And influencing each other between gene, in the urgent need to a kind of engineering that in vivo can quickly knock out and insert gene. CRISPR/Cas9 system occurs at once, meets this demand of researcher.

CRISPR/Cas9 system (Clustered regularly interspaced short palindromic Repeats/CRISPR-associated protein 9) it is a kind of simple to operate, applicability extensive gene editing instrument. Whole system is mainly risen by a nucleic acid cleaving enzymatic (Cas9) and one and guides the RNA (sgRNA) of recognition reaction to form.sgRNA Identified with target gene site by base pair complementarity, then recruit Cas9 and be digested, produce double-strand break, thus realize The gene editing of DNA level.It because its applicability is wide, convenient and time-saving, is quickly applied to various aspects, especially builds at cancer model Vertical and gene therapy probe into aspect, have very big superiority.

But, discovery in the constantly exploration of scientist, the different sgRNA's designing for same gene in same cell Digesting efficiency has very big difference, if high efficiency sgRNA can not be designed, can only be made up by increasing concentration, so will Can a lot of gene rubbish be brought to cell, produce a high proportion of missing the target simultaneously, give scientific research personnel research bring very big not Just, the sgRNA therefore designing a high digesting efficiency is extremely important for the research in terms of gene.

At present, the design software of existing sgRNA has nearly 30 kinds, is broadly divided into two classes：One class is to sum up from experiment Some of sgRNA are regular, and PAM sequence must be contained in the sgRNA sequence one end for example matched, and 5 ' ends should be GG, G/C content Should be maintained at about 60%, Seed Sequences can't stand mispairing etc., is then directly screened by arranging condition,；Another kind of master To calculate the specific of sgRNA by using statistical method to give a weight to each base, such as CRISPR Design.What the software of both types was all set up is the model of a versatility, yet with different plant species and different cell Between have very big heterogeneity, cause the prediction efficiency of existing software to be not very well, and because different under different experimental conditions The digesting efficiency to sgRNA for the matter has a certain impact, and general model evaluation accuracy rate ratio is relatively low.

Accordingly, it is considered to the heterogeneity between different platform species data, set up individual by the data of different platform or species Property model to improve the specific of sgRNA and high efficiency, extremely heavy for the miss the target research of problem of CRISPR/Cas9 system Want.

Content of the invention

It is an object of the invention to for the problems referred to above provide that a kind of accuracy rate is high, applied range based on CRISPR/ The method for designing of the sgRNA of Cas9.

For realizing purpose of the present invention, the present invention provides the method for designing of a kind of sgRNA based on CRISPR/Cas9, The method comprises the following steps：

1) obtain the value of the digesting efficiency of sgRNA and corresponding Cas9, be specially：

11) from document, obtain the value of the digesting efficiency of sgRNA and corresponding Cas9；

12) from SRA database, obtain sgRNA, calculate the value of the digesting efficiency obtaining corresponding Cas9；

13) according to species, cell type and experiment condition by step 11) and 12) in the data that get be categorized into difference Reference genome, each with reference to genome is listed a first be classified as sgRNA title, second be classified as sgRNA sequence with And the 3rd form of digesting efficiency being classified as corresponding Cas9；

2) set up personalized sgRNA to design a model, be specially：

21) according to demand from corresponding with reference to genome, extraction step 1) in the sequence information of sgRNA that obtains；

22) to step 21) in extract sgRNA sequence information carry out binary coding according to binary rules；

23) to step 21) the middle sgRNA obtaining, it is judged that the data type of the digesting efficiency of its Cas9, if numeric type is then Enter step 24), if classifying type then enters step 25)；

24) to step 22) in coding after sgRNA sequence information, carry out feature extraction with Lasso model, according to standard Linear regression is set up personalized sgRNA and is designed a model；

25) to step 22) in coding after sgRNA sequence information, with two sorted logics recurrence in L1 regularization carry out Feature selecting, the L2 regularization in returning further according to two sorted logics is set up personalized sgRNA and is designed a model；

3) use NDCG algorithm to weigh step 2) in the quality that designs a model of personalized sgRNA set up update SRA number It according to storehouse, is specially：

31) calculation procedure 2) in the NDCG value that designs a model of personalized sgRNA set up；

32) judge whether existing SRA database has corresponding personalized sgRNA model, if being otherwise added to SRA Database, if then entering step 33)；

33) compare this personalization sgRNA model and the sgRNA model in corresponding SRA database, select NDCG value big One is stored in SRA database；

4) design sgRNA the assessed value providing each sgRNA, be specially：

41) genome area being given according to user, it is suitable with reference to genome to choose from SRA database, therefrom searches Suo Suoyou meets the sgRNA of design rule, as the sgRNA of design；

42) to step 41) in design sgRNA, use step 2) in set up personalized sgRNA model be estimated.

Preferably, described step 12) in be calculated corresponding Cas9 the value of digesting efficiency be specially：

121) the long comparison of reading of sgRNA and corresponding two generations order-checking to reference on genome；

122) reading comprising sgRNA is taken out long；

123) judge cut point whether produce on DNA insert or whether the insertion on deletion and DNA or deletion are Frameshift mutation；

124) add up the frameshift mutation rate of each sgRNA, be specially：

125) using step 124) in calculated frameshift mutation rate as the value of the digesting efficiency of Cas9.

Preferably, described step 21) in the sequence information of sgRNA include that sgRNA sequence, sgRNA identify the required mark of DNA The base of the upstream and downstream of the spacer of master chip section and sgRNA, the bases longs of the upstream and downstream of the spacer of described sgRNA is flat Platform default value or the value of user setup.

Preferably, described step 22) in binary rules be specially：A correspondence 1000, C correspondence 0100, G correspondence 0010, T correspondence 0001, N correspondence 0000.

Preferably, described step 24) in Lasso model carry out feature extraction be by extract non-zero weight select spy Levy vector, be specially：

Wherein, w is the weight of estimative characteristic vector, and x is the characteristic vector of selected sgRNA, and n is sgRNA Quantity, y is the value of the digesting efficiency of the corresponding Cas9 of sgRNA；α is a constant, | | w | |₁It is the matrix of parameter vector； Lasso model is by increasing α | | w | |₁Solve this least square loss function, by traversal regularization matrix, non-zero weight Feature be extracted.

Preferably, described step 25) in L1 regularization be specially：

Wherein, w and c is weight and the intercept of estimative feature, and X is the binary matrix of the sgRNA of coding, and n is The quantity of sgRNA, y is the value of the digesting efficiency of the corresponding Cas9 of sgRNA.

Preferably, described L2 regularization is specially：

Preferably, described step 31) in calculate the NDCG value that designs a model of personalized sgRNA set up and be specially：

Wherein, DCG is the numerical value calculating with prediction sequence, and IDCG is the preferable DCG calculating gained with true sequence, rel_iIt is the ranking value of the i-th position prediction.

Preferably, described step 41) in design rule be specially：

20bp+PAM

Wherein, bp is for representing the unit of DNA length, and PAM is that sgRNA identifies the required mark fragment of DNA.

Compared with prior art, the invention have the advantages that：

(1) for the dissimilar cell of different plant species, the strategy of personalization is employed, and with the machine learning of data-driven Algorithm is modeled, and assessment accuracy rate has significant improvement.

(2) new coding rule is used so that the feature finding is more complete, is not limited only between PAM and spacer.

(3) impart user oneself and build the flow process of model so that range of application is wider, be not limited only in database only have Some species.

(4) use the OTF of NGS data to lead as being digested rate, expand the scope that can analyze data；

(5) user can upload the data of oneself and comes expanding data storehouse, accelerates the accumulation of data, advantageously accounts for now Factor data amount deficiency causes the predicament that can not design optimum sgRNA very well.

Brief description

Fig. 1 is for setting up the method flow diagram of personalized sgRNA model and model evaluation；

Fig. 2 is the method flow diagram of design and assessment sgRNA.

Detailed description of the invention

The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.The present embodiment is with technical solution of the present invention Premised on implement, give detailed embodiment and concrete operating process, but protection scope of the present invention be not limited to Following embodiment.

Abbreviation explanation：

CRISPR：Clustered regularly interspaced short palindromic repeats

The little palindrome repetitive sequence at the interval of the rule of cluster

Cas9：With the related enzyme of CRISPR II type system

NGS：Next Generation Sequencing, the order-checking of two generations

PAM：Protospacer-adjacent motif, sgRNA identify the required mark fragment of DNA

sgRNA：CRISPR/Cas9 system rises the RNA of guiding function

indel：CRISPR/Cas9 edits the insertion on the DNA causing, deletion

spacer：SgRNA rises the base of about 20 of base pair complementarity

OTF：Out of frame, frameshift mutation.

Read：Read long, be the sequencing sequence that in high-flux sequence, a reaction obtains.

The present embodiment provides the method for designing of a kind of sgRNA based on CRISPR/Cas9, dissimilar for different plant species Cell sets up the flow process that oneself personalized sgRNA designs a model, and can set up model according to different demands and design sgRNA, tool Body includes following four step：

(1) Data Collection：The receipt collected from document is generally two classes：SgRNA and corresponding digesting efficiency number Value type or sgRNA and corresponding digesting efficiency classifying type (such as effective or invalid two classification)；Download from SRA database NGS then only have numeric type a kind of.Because NGS data are by the numeric type one collected in flow process and the document after adding up OTF and leading Cause, therefore document classification type and two kinds of data of NGS are only illustrated by the present embodiment.

Classifying type data：For the classifying type data collected from document, the present embodiment regulation is effectively 1, and invalid is 0, It is organized into such as the form of table 1.

Table 1

sgID	Sequence	Score
			sgRNA_1	CGCAACCTGCTCAGCGCCTACGG	1
sgRNA_2	CAGTCTACATAACACGCCCATGG	1
			sgRNA_3	CGCAACCTGCTCAGCGCCTACGG	1
……	……	……
			sgRNA_1_1	GGCAACCGTGGCGGCAATCGAGG	0
sgRNA_2_2	CTTCTCGGAATTCGGTGAAGGTGG	0
			sgRNA_3_3	AACCTCCCGGCTTCTCGGAATTCGG	0
……	……	……

Numeric type data：For the numeric type data of NGS, first pass through BWA respectively the sequence of sgRNA and NGS Reads comparison with reference on genome to the mankind, is taken out the reads comprising sgRNA, and is judged whether cut point produces indel And whether indel is OTF, then adds up the OTF of each sgRNA and lead that (OTF leads=comprises this sgRNA and is OTF The sum of reads is divided by the total reads number comprising this sgRNA).Finally arrange as such as the form of table 2.

Table 2

sgID	Sequence	Score
			sgRNA_1	CGCAACCTGCTCAGCGCCTACGG	0.2345
sgRNA_2	CAGTCTACATAACACGCCCATGG	0.7846
			sgRNA_3	CGCAACCTGCTCAGCGCCTACGG	0.2367
……	……	……

(2) model is set up：As it is shown in figure 1, from the corresponding sequence information extracting the sgRNA collecting with reference to genome. Assume that arranging upstream and downstream sequence is respectively 35 and 32 bases, then the sequence taken out is 90 (35+20+3+32) individual base. CACCTGGTAT GTTCGTATCG GGCAGAATATCGCAACCTGC TCAGCGCC TA CGGTCCATCT CGCTCAGGTA CGACTGACCGACCCAGTCTA.

Carry out binary coding to the sgRNA information extracted, rule is as shown in table 3.

Table 3

Then 90 bases of above taking-up may be encoded as：

0100 1000 0100 0100 0001 0010 0010 0001 1000 0001

0010 0001 0001 0100 0010 0001 1000 0001 0100 0010

0010 0010 0100 1000 0010 1000 1000 0001 1000 0001

0100 0010 0100 1000 1000 0100 0100 0001 0010 0100

0001 0100 1000 0010 0100 0010 0100 0100 0001 1000

0100 0010 0010 0001 0100 0100 1000 0001 0100 0001

0100 0010 0100 0001 0100 1000 0010 0010 0001 1000

0100 0010 1000 0100 0001 0010 1000 0100 0100 0010

1000 0100 0100 0100 1000 0010 0001 0100 0001 1000

Extract feature with machine learning method, set up personalized sgRNA and design a model.

For classifying type data, select feature with logistic regression and set up forecast model.Two sorted logics return two Individual optional regularization, present invention L1 regularization carries out feature selecting, and model is set up in L2 regularization.

The optimization problem that the following sparse features of L1 regularization logistic regression solution selects：

Wherein, w and c is weight and the intercept of estimative feature, and X is the character representation of training sample, and n is training sample Quantity, y is the corresponding digesting efficiency value of sgRNA.

Minimize cost function by L2 punishment logistic regression solution：

For numeric type data, doing feature selecting with Lasso model, standard linear regression sets up forecast model. Lasso is the linear model estimating sparse coefficient correlation, mainly selects characteristic vector by extracting non-zero weight.Minimize mesh Scalar functions is：

Wherein, w is the weight of estimative characteristic vector, and x is the characteristic vector of selected sgRNA, and n is training sample Quantity, y is the corresponding digesting efficiency value of sgRNA；α is a constant, | | w | |₁It is the matrix of parameter vector；Lasso mould Type is by increasing α | | w | |₁Solve this least square loss function, by traversal regularization matrix, the feature quilt of non-zero weight Extracting, these features are considered as the important element affecting sgRNA digesting efficiency.

After choosing these features, then set up an assessment models with a standard linear regression.

The modeling result of numeric type and classifying type all produces two files：One is xml document, and content comprises selectively Feature, and the result of cross validation；Another file is pkl file, and content is the forecast model set up, binary file.

Xml document content is as follows：

(3) assessment models：NDCG algorithm is used to weigh the quality of forecast model, NDCG (Normalized Discounted Cumulative Gain, storage gain is lost in normalization) it is the usefulness being mainly used to weigh an order models, its value generation The similitude between the ranking results of prediction and the sequence of reality by table, and between zero and one, 1 represents completely the same to scope, numerical value Represent more greatly this model better.Concrete formula is as follows：

DCG (Discounted Cumulative Gain, lose storage gain) is the numerical value calculating with prediction sequence, IDCG (ideal DCG), is preferable DCG, calculates gained with true sequence.The mathematical definition of DCG is as follows：

Wherein, rel_iIt is the ranking value of the i-th position prediction.

As shown in the table, sgID is the title of sgRNA, and seq is the spacer sequence of sgRNA, and Benchmark Score is Benchmark fraction, BS_rank is the sequence of Benchmark Score, and Cage is the fraction of forecast model of the present invention assessment, C_rank Sequence for Cage is as shown in table 4.

Table 4

sgID	seq	Benchmark Score	BS_rank	Cage Score	C_rank
						sg1000	GCAGGTACCCTGCAACGTCGCGG	0.789456865	1	0.6905	1
sg1001	CTCCACTAGTCCCCGCGCCGCGG	0.506422166	2	0.6026	2
						sg1	GTAATGGCTTCCTCGTGAGTTGG	0.325738326	3	0.5548	3
sg1002	GACTCCGTTGGGATCCGCGCCGG	0.092078991	4	0.5095	4
						sg10	ATCTTAAGCAAACGCTTACCAGG	0.072255575	5	0.4959	5
sg1003	CCCGAAACGGTTGACTCCGTTGG	0.037552375	6	0.4473	6
						sg1004	AGGCGCGCGATCCAGGTAGCTGG	0.019922477	7	0.3281	8
sg100	AAAAAGCTGATGAAGTTGTTTGG	0.017296539	8	0.3357	7
						sg1005	CGGGGCCACCGCGACGTTGCAGG	0.002206787	9	0.3056	9
……	……	……	……	……	……

TOP50 NDCG=0.876322904

TOP 10%NDCG=0.84340749

If not having this model in database, then updating and arriving database, the NDCG value otherwise calculating two groups compares, if New model is bigger than the NDCG value of existing model, then renewable to database.

(4) design and assessment：As in figure 2 it is shown, be estimated for the sgRNA that user has designed or for user to The genome area (such as chromosome 1,1,000,000to 1,002,000, hg19) going out, carries out the design of sgRNA, first First determine species or the cell type of sgRNA to be assessed, then select the model being suitable for be estimated, without properly Model, optional similar model, present embodiments provide that to relate to 10 models of 8 kinds of cells of 3 species for you to choose Use.Result output is as shown in table 5.

Table 5

So far, user can select the sgRNA being suitable for oneself demand to carry out next step research.

Claims (9)

1. the method for designing based on the sgRNA of CRISPR/Cas9, it is characterised in that the method comprises the following steps：

1) obtain the value of the digesting efficiency of sgRNA and corresponding Cas9, be specially：

11) from document, obtain the value of the digesting efficiency of sgRNA and corresponding Cas9；

12) from SRA database, obtain sgRNA, calculate the value of the digesting efficiency obtaining corresponding Cas9；

13) according to species, cell type and experiment condition by step 11) and 12) in the data that get be categorized into different ginsengs Examining genome, each a first is classified as sgRNA title, second is classified as sgRNA sequence and the with reference to listing in genome The form of three digesting efficiency being classified as corresponding Cas9；

2) set up personalized sgRNA to design a model, be specially：

21) according to demand from corresponding with reference to genome, extraction step 1) in the sequence information of sgRNA that obtains；

22) to step 21) in extract sgRNA sequence information carry out binary coding according to binary rules；

23) to step 21) the middle sgRNA obtaining, it is judged that the data type of the digesting efficiency of its Cas9, if numeric type then enters Step 24), if classifying type then enters step 25)；

24) to step 22) in coding after sgRNA sequence information, carry out feature extraction with Lasso model, according to normal linearity Return the personalized sgRNA of foundation to design a model；

25) to step 22) in coding after sgRNA sequence information, with two sorted logics recurrence in L1 regularization carry out feature Selecting, the L2 regularization in returning further according to two sorted logics is set up personalized sgRNA and is designed a model；

3) use NDCG algorithm to weigh step 2) in the quality that designs a model of personalized sgRNA set up update SRA database, It is specially：

31) calculation procedure 2) in the NDCG value that designs a model of personalized sgRNA set up；

32) judge whether existing SRA database has corresponding personalized sgRNA model, if being otherwise added to SRA data Storehouse, if then entering step 33)；

33) the sgRNA model in this personalization sgRNA model and corresponding SRA database is compared, select that NDCG value is big one It is stored in SRA database；

4) design sgRNA the assessed value providing each sgRNA, be specially：

41) genome area being given according to user, it is suitable with reference to genome to choose from SRA database, therefrom searches for institute There is the sgRNA meeting design rule, as the sgRNA of design；

42) to step 41) in design sgRNA, use step 2) in set up personalized sgRNA model be estimated.

2. the method for designing of the sgRNA based on CRISPR/Cas9 according to claim 1, it is characterised in that described step 12) value of the digesting efficiency being calculated corresponding Cas9 in is specially：

121) the long comparison of reading of sgRNA and corresponding two generations order-checking to reference on genome；

122) reading comprising sgRNA is taken out long；

123) judge whether the insertion in the insertion whether cut point produces on DNA or deletion and DNA or deletion are frameshit Sudden change；

124) add up the frameshift mutation rate of each sgRNA, be specially：

125) using step 124) in calculated frameshift mutation rate as the value of the digesting efficiency of Cas9.

3. the method for designing of the sgRNA based on CRISPR/Cas9 according to claim 1, it is characterised in that described step 21) in, the sequence information of sgRNA includes that sgRNA sequence, sgRNA identify the spacer of the required mark fragment of DNA and sgRNA The base of upstream and downstream, the bases longs of the upstream and downstream of the spacer of described sgRNA is the value of platform default value or user setup.

4. the method for designing of the sgRNA based on CRISPR/Cas9 according to claim 1, it is characterised in that described step 22) binary rules in is specially：A correspondence 1000, C correspondence 0100, G correspondence 0010, T correspondence 0001, N correspondence 0000.

5. the method for designing of the sgRNA based on CRISPR/Cas9 according to claim 1, it is characterised in that described step 24) carrying out feature extraction with Lasso model in is to select characteristic vector by extracting non-zero weight, is specially：

$\underset{w}{min}\frac{1}{2n}\left|\right|xw-y|{|}_2^2+\mathrm{\α}|\left|w\right|{|}_1$

Wherein, w is the weight of estimative characteristic vector, and x is the characteristic vector of selected sgRNA, and n is the quantity of sgRNA, Y is the value of the digesting efficiency of the corresponding Cas9 of sgRNA；α is a constant, | | w | |₁It is the matrix of parameter vector；Lasso model By increasing α | | w | |₁Solving this least square loss function, by traversal regularization matrix, the feature of non-zero weight is carried Take out.

6. the method for designing of the sgRNA based on CRISPR/Cas9 according to claim 1, it is characterised in that described step 25) the L1 regularization in is specially：

$\underset{w,c}{min}\frac{1}{2}\left|\right|w|{|}_1+{\mathrm{C\Σ}}_{i=1}^{n}log\left(\exp \right(-y_i\left(X_i^{T}w+c\right))+1)$

Wherein, w and c is weight and the intercept of estimative feature, and X is the binary matrix of the sgRNA of coding, and n is sgRNA Quantity, y is the value of the digesting efficiency of the corresponding Cas9 of sgRNA.

7. the method for designing of the sgRNA based on CRISPR/Cas9 according to claim 6, it is characterised in that described L2 is just Then change and be specially：

$\underset{w,c}{min}\frac{1}{2}{w}^{T}w+{\mathrm{C\Σ}}_{i=1}^{n}log\left(\exp \right(-y_i\left(X_i^{T}w+c\right))+1).$

8. the method for designing of the sgRNA based on CRISPR/Cas9 according to claim 1, it is characterised in that described step 31) the NDCG value that the personalized sgRNA that in, calculating is set up designs a model is specially：

$NDCG=\frac{DCG}{IDCG}$

$DCG={\mathrm{rel}}_1+\underset{i=2}{\overset{n}{\Σ}}\frac{{\mathrm{rel}}_i}{{\log }_{2}i}$

Wherein, DCG is the numerical value calculating with prediction sequence, and IDCG is the preferable DCG, rel calculating gained with true sequence_iIt is The ranking value of the i-th position prediction.

9. the method for designing of the sgRNA based on CRISPR/Cas9 according to claim 1, it is characterised in that described step 41) in, design rule is specially：

20bp+PAM

Wherein, bp is for representing the unit of DNA length, and PAM is that sgRNA identifies the required mark fragment of DNA.