Introduction

Linkage disequilibrium (LD) is the association for a pair of loci and the metric of LD serves as the basis for developing genetic applications in agriculture, evolutionary biology, and biomedical researches (Weir, 2008; Hill and Robertson, 1966). The structure of LD of the human genome is shaped by many factors, mutation, recombination, population demography, epistatic fitness, and completeness of genomic data itself (Myers et al., 2005; Nei and Li, 1973; Ardlie et al., 2002). Due to its overwhelming cost, LD structure investigation is often compromised to a small genomic region (Chang et al., 2015; Theodoris et al., 2021), and their typical LD structure is as illustrated for a small segment (Barrett et al., 2005). Now, given the availability of large-scale genomic data, such as millions of single nucleotide polymorphisms (SNPs), the large-scale LD patterns of the human genome play crucial roles in determining genomics studies, and many theories and useful algorithms upon large-scale LD structure, from genome-wide association studies, polygenic risk prediction for complex diseases, and choice for reference panels for genotype imputation (Vilhjálmsson et al., 2015; Yang and Zhou, 2020; Bulik-Sullivan et al., 2015; Yang et al., 2011; Das et al., 2016).

However, there are impediments, largely due to intensified computational cost, in both investigating large-scale LD and providing high-resolution illustration for their details. If we consider a genomic grid that is consisted of m² SNP pairs, given a sample of n individuals and m SNPs (n «m) – typically as observed in 1000 Genomes Project (1KG) (Lowy-Gallego et al., 2019), its benchmark computational time cost for calculating all pairwise LD is 𝒪 (nm²), a burden that quickly drains computational resources given the volume of the genomic data. In practice, it is of interest to know the mean LD of the SNP pairs for a genomic grid, which covers m_i ×m_j SNP pairs. Upon how a genomic grid is defined, a genomic grid consequently can be consisted of : i) the whole genome-wide m² SNP pairs, and we denote their mean LD as ℓg ; ii) the intra-chromosomal mean LD for the ith chromosome of SNP pairs, and denote as ℓi ; iii) the inter-chromosomal mean LD and chromosomal m_i m_j SNP pairs, and denoted as ℓi· j .

In this study we propose an efficient algorithm that can estimate ℓg, ℓi, and ℓ i· j ·, the computational time of which can be reduced from to 𝒪 (n²m_i) for ℓi and 𝒪 (nm_i m_j) to 𝒪 (n²m_i+n²m_j ) for ℓi· j. The rationale of the proposed method relies on the connection between the genetic relationship matrix (GRM) and LD (Chen, 2014; Goddard, 2009), and in this study a more general transformation from GRM to LD can be established via Isserlis’s theorem (Isserlis, 1918; Zhou, 2017). The statistical properties, such as sampling variance, of the estimated LD have been derived too.

The proposed method can be analogously considered a more powerful realization for Haploview (Barrett et al., 2005), but additional utility can be derived to bring out unprecedented survey of LD patterns of the human genome. As demonstrated in 1KG, we consequently investigate how biological factors such as population structure, admixture, or variable local recombination rates can shape large-scale LD patterns of the human genomes.

1. The proposed method provides statistically unbiased estimates for large-scale LD patterns and shows computational merits compared with the conventional methods (Figure 2).

2. We estimated ℓg, and 22 autosomal ℓi. and 231 inter-autosomal ℓi· j for the 1KG cohorts. There were ubiquitously existence of extended LD, which was associated with population structure or admixture (Figure 3).

3. We provided high-resolution illustration that decomposed a chromosome into upmost nearly a million grids, each of which was consisted of 250×250 SNP pairs, the highest resolution that has been realized so far at autosomal level (Figure 4); tremendous variable recombination rates led to regional strong LD as highlighted for the HLA region of chromosomes 6 and the centromere region of chromosome 11.

4. Furthermore, a consequently linear regression constructed could quantify LD decay score genome-widely, and in contrast LD decay was previously surrogated in a computational expensive method. There was strong ethnicity effect that was associated with extended LD (Figure 5).

5. We demonstrate that the strength of autosomal ℓi. was inversely proportional to the SNP number, an anticipated relationship that is consistent to genome-wide spread of recombination hotspots. However, the chromosome 8 of ASW showed substantial deviation from the fitted linear relationship (Figure 6).

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

The proposed algorithm has been realized in C++ and is available at: https://github.com/gc5k/gear2. As tested the software could handle sample size as large as 10,000 individuals.

Methods and Materials

The overall rationale for large-scale LD analysis

We assume LD for a pair of biallelic loci is measured by squared Pearson’s correlation, in which the LD of loci l₁ and l₂, p and q the reference and the alternative allele frequencies. If we consider the averaged LD for a genomic grid over SNP pairs, the conventional estimator is and, if we consider the averaged LD for m_i and m_j SNP pairs between two genomic segments, then . Now let us consider the 22 human autosomes (Figure 1A). We naturally partition the genome into 𝒞=22 blocks, and its genomic LD, denoted as ℓg can be expressed as

So we can decompose ℓg into 𝒞 ℓi. and unique ℓi· j . Obviously, Eq 1 can be also expressed in the context for a single chromosome , in which the number of SNP segments, each of which has m SNPs. Geometrically it leads to β_i diagonal grids and unique off-diagonal grids (Figure 1B).

LD-decay regression

As human genome can be boiled down to small LD blocks by genome-widely spread recombination hotspots (Hinch et al., 2019; Li et al., 2022), mechanically there is self-similarity for each chromosome that the relatively strong ℓi. for juxtaposed grids along the diagonal but weak ℓi· j for grids slightly off-diagonal. So, for a chromosomal ℓi., we can further express it as

in which ℓu is the mean LD for a diagonal grid, ℓuv the mean LD for off-diagonal grids, and m_i the number of SNPs on the i^th chromosome. Consider a linear model below (Figure 1C for its illustration),

in which ℓ represents a vector composed of 𝒞 ℓi., x represents a vector composed of 𝒞 x_i and the inversion of the SNP number of the chromosome. The regression coefficient and intercept can be estimated as below

And

There are some technical details in order to find the interpretation for b₀ and b₁ .We itemize them briefly. For the mean and variance of x:

For E(xℓ)

For E(x) For E(ℓ)

Then we integrate these items to have the expectation for b₁:

d

Similarly, we plug in E(b₁) so as to derive b₀ :

After some algebra, if E(ℓ_u) ≫ E (ℓ_uv) – say if the former is far greater than the latter, the interpretation of b₁ and b₀ can be

It should be noticed that E(b₁) ≈ E(ℓ_u)m quantifies the averaged LD decay of the genome. Conventional LD decay is analysed via the well-known LD decay analysis, but Eq 4 provides a direct estimate of both LD decay and possible existence of extended LD. We will see the application of the model in Figure 5 that the strength of the long-distance LD is associated with population structure. of note, the underlying assumption of Eq 3 and Eq 4 is genome-wide spread of recombination hotspots, an established result that has been revealed and confirmed (Hinch et al., 2019).

Efficient estimation for ℓ_g, ℓ_i, and ℓ_{i ·j}

For the aforementioned analyses, the bottleneck obviously lies in the computational cost in estimating ℓ_i and ℓ_i·j· ℓ_i and ℓ_i·j are used to be estimated via the current benchmark algorithm as implemented in PLINK (Chang et al., 2015), and the computational time complex is proportional to O(nm²). We present a novel approach to estimate ℓ_i and ℓ_i·j· Given a genotypic matrix X, a n × m mmatrix, if we assume that there are m_i and m_j SNPs on chromosomes i and j, respectively, we can construct n × n genetic relatedness matrices as below

in which is the standardized X_i and where x_kl is the genotype for the K^th individual at the l^th biallelic locus, F is the inbreeding coefficient having the value of 0 for random mating population and 1 for an inbred population, p_l and q_l are the frequencies of the reference and the alternative alleles (p_l + q_l =1)respectively. When GRM is given, we can obtain some statistical characters of K_i. We extract two vectors , which stacks the off-diagonal elements of K_i, and which takes the diagonal elements of K_i. The mathematical expectation of in which can be established according to Isserlis’s theorem in terms of the four-order moment (Isserlis, 1918),

in which is the expected relatedness score and r indicates the r^th degree relatives. r=0 for the same individual, and r=1 for the first degree relatives. Similarly, we can derive for Eq 6 establishes the connection between GRM and the aggregated LD estimation that .According to Delta method as exampled in Appendix I of Lynch and Walsh’s book (Lynch and Walsh, 1998), the means and the sampling variances for ℓ_i and ℓ_i·j are,

in which and , respectively. Of note, the properties of ℓ_g can be derived similarly if we replace ℓ_i with ℓ_g in Eq 7. We can develop , a scaled version of ℓ_i·j· as below

in which , a modification that removed the LD with itself. According to Delta method, the sampling variance of is

Of note, when there is no LD between a pair of loci, ℓ yields zero and its counterpart PLINK estimate yields ,a difference that can be reconciled in practice (see Figure 2).

Raise of LD due to population structure

In this study, the connection between LD and population structure is bridged via two pathways below, in terms of a pair of loci and of the aggregated LD for all pair of loci. For a pair of loci, their LD is often simplified as , but will be inflated if there are subgroups (Nei and Li, 1973). In addition, it is well established the connection between population structure and eigenvalues, and in particular the largest eigenvalue is associated with divergence of subgroups (Patterson et al., 2006). In this study, the existence of subgroups of cohort is surrogated by the largest eigenvalue .

Data description and quality control

The 1KG (Auton et al., 2015), which is launched to produce a deep catalogue of human genomic variation by whole genome sequencing (WGS) or whole exome sequencing (WES), and 2,503 strategically selected individuals of global diversity are included (containing 26 cohorts). We used the following criteria for SNP inclusion for each of the 26 1KG cohorts: i) autosomal SNPs only; ii) SNPs with missing genotype rates higher than 0.2 were removed, and missing genotypes were imputed; iii) Only SNPs with minor allele frequencies higher than 0·05 were retained. Then 2,997,635 consensus SNPs that were present in each of the 26 cohorts were retained. According to their origins, the 26 cohorts are grouped as African (AFR: MSL, GWD, YRI, ESN, ACB, LWK, and ASW), European (EUR: TSI, IBS, CEU, GBR, and FIN), East Asian (EA: CHS, CDX, KHV, CHB, and JPT), South Asian (SA: BEB, ITU, STU, PJL, and GIH), and American (AMR: MXL, PUR, CLM, and PEL), respectively.

In addition, to test the capacity of the developed software (X-LD), we also included CONVERGE cohort (n=10,640), which was used to investigate Major Depressive Disorder (MDD) in the Han Chinese population (Cai et al., 2015). We performed the same criteria for SNP inclusion as that of the 1KG cohorts, and m=5,215,820 SNPs were remained for analyses.

X-LD software implementation

The proposed algorithm has been realized in our X-LD software, which is written in C++ and reads in binary genotype data as often used in PLINK. As multi-thread programming is adopted, the efficiency of X-LD can be improved upon the availability of computational resources. We have tested X-LD in various independent datasets for its reliability and robustness. Certain data management options, such as flexible inclusion or exclusion of chromosomes, have been built into the commands of X-LD. In X-LD, missing genotypes are naively imputed according to Hardy-Weinberg proportions; however, when missing rate is high, we suggest the genotype matrix should be imputed by other advanced imputation tools.

The most time-consuming part of X-LD was the construction of GRM , and the established computational time complex was 𝒪 (n²m). However, if is decomposed into , in which , has dimension of n × B, using Mailman algorithm the computational time complex for building K can be reduced to (Liberty and Zucker, 2009). This idea of embedding Mailman algorithm into certain high throughput genomic studies has been successful, and our X-LD software is also leveraged by absorbing its recent practice in genetic application (Wu and Sankararaman, 2018).

Results

Statistical properties of the proposed method

Table 1 introduced symbols frequently cited in this study. As schematically illustrated in Figure 1, ℓ_g could be decomposed into 𝒞 ℓ_i and unique ℓ_i,j components. We compared the estimated ℓ;_i and ℓ_i,j in X-LD with those being estimated in PLINK (known as “--r2”, and the estimated squared Pearson’s correlation LD is denoted as r²). Considering the 212 substantial computational cost of PLINK, only 100,000 randomly selected autosome SNPs were used for each 1KG cohort, and 22 and 231 were estimated. After regressing 22 against those of PLINK, we found that the regression slope was close to unity and bore an anticipated intercept a quantity of approximately (Figure 2A and Figure 2B).

Table 1.

In other words, PLINK gave even for SNPs of no LD. However, when regressing 231 estimates against those of PLINK, it was found that largely because of tiny quantity of it was slightly smaller than 1 but statistically insignificant from 1 in these 26 1KG cohorts (mean of 0.86 and s.d. of 0.10, and its 95 % confidence interval was (0.664, 1.056)); when the entire 1KG samples were used, its much larger LD due to subgroups, nearly no estimation bias was found (Figure 2A and Figure 2B). In contrast, because of their much larger values, components were always consistent with their corresponding estimates from PLINK (mean of 1.03 and s.d. of 0.012, 95% confidence interval was (1.006, 1.053), bearing an ignorable bias). Furthermore, we also combined the African cohorts together (MSL, GWD, YRI, ESN, LWK, totaling 599 individuals), the East Asian cohorts together (CHS, CDX, KHV, CHB, and JPT, totaling 504 individuals), and the European cohorts together (TSI, IBS, CEU, GBR, and FIN, totaling 503 individuals), the resemblance pattern between X-LD and PLINK was similar as observed in each cohort alone (Figure S1). The empirical data in 1KG verified that the proposed method was sufficiently accurate.

To fairly evaluate the computational efficiency of the proposed method, the benchmark comparison was conducted on the first chromosome of the entire 1KG dataset (n =2.503) and m =(225,967), and 10 CPUs were used for multi-thread computing. Compared with PLINK, the calculation efficiency of X-LD was nearly 30∼40 times faster for the tested chromosome, and its computational time of X-LD was proportional to (Figure S2). So, X-LD provided a feasible and reliable estimation of large-scale complex LD patterns. More detailed computational time of the tested tasks would be reported in their corresponding sections below; since each 1KG cohort had sample size around 100, otherwise specified the computational time was reported for CHB (n = 103) as a reference (Table 2). In order to test the capability of the software, the largest dataset tested was CONVERGE (n =10,640, and (n =5,215,820), and it took 77,508.00 seconds, about 22 hours, to estimate 22 autosomal and 231 (Figure 1A);

Table 2.

When zooming into chromosome 2 of CONVERGE, on which 420,949 SNP had been evenly split into 1,000 blocks and yielded 1,000 grids, and 499,500 LD grids, it took 45,125.00 seconds, about 12.6 hours, to finished the task (Figure 1B).

Ubiquitously extended LD and population structure/admixture

We partitioned the 2,997,635 SNPs into 22 autosomes (Figure 3A and Figure S3), and the general LD patterns were as illustrated for CEU, CHB, YRI, ASW, and 1KG. As expected, for each cohort (Figure 3B). As observed in these 1KG cohorts, all these three LD measures were associated with population structure, which was surrogated by , and their squared correlation R² were greater than 0.8. ACB, ASW, PEL, and MXL, which all showed certain admixture, tended to have much greater , and (Table 3 and Figure 3B). In contrast, East Asian (EA) and European (EUR) orientated cohorts, which showed little within cohort genetic differentiation – as their largest eigenvalues were slightly greater than 1, had their aggregated LD relatively low and resembled each other (Table 3). Furthermore, for several European (TSI, IBS, and FIN) and East Asian (JPT) cohorts, the ratio between and components could be smaller than 0.1, and the smallest ratio was found to be about 0.091 in FIN. The largest ratio was found in 1KG that and , and the ratio was 0.877 because of the inflated LD due to population structure. A more concise statistic to describe the ratio between ℓ_i,jand ℓ;_i was , Eq 8, and the corresponding values for 231 scaled for FIN was (s.d. of 0.027) and for 1KG was (s.d. of 0.028).

Table 3.

In terms of computational time, for 103 CHB samples, it took about 101.34 seconds to estimate 22 autosomal and 231 ; for all 1KG 2,503 samples, X-LD took about 3,008.29 seconds (Table 1). Conventional methods took too long to complete the analyses in this section, so no comparable computational time was provided. For detailed 22 and 231 estimates for each 1KG cohort, please refer to Extended Data 1 (Excel Sheet 1-27).

Detecting exceedingly high LD grids shaped by variable recombination rates

We further explored each autosome with high-resolution grid LD visualization. We set m =250, so each grid had the ℓ;_uv for 250 × 250 SNP pairs. The computational time complex was , in which , and with our proposed method in CHB it costed 66.86 seconds for chromosome 2, which had 241,241 SNPs and was totaled 466,095 unique grids, and 3.22 seconds only for chromosome 22, which had 40,378 SNPs and was totaled 13,203 unique grids (Table 1). In contrast, under conventional methods those LD grids were not very likely to be exhaustively surveyed because of its computational cost was : for CHB chromosome 2, it would have taken about 40 hours as estimated. As the result was very similar for m =500 (Figure S4), we only reported the results under m =250 below.

As expected, chromosome 6 (206,165 SNPs, totaling 340,725 unique grids) had its HLA cluster showing much higher LD than the rest of chromosome 6. In addition, we found very dramatic variation of the HLA cluster LD HLA (28,477,797-33,448,354 bp, totaling 3,160 unique grids) across ethnicities. For CEU, CHB, YRI, and ASW, their , and 0.0019, respectively, but their corresponding HLA cluster grids had and 0.022, respectively (Figure 4).

Consequently, the largest ratio for was of 42.00 in CEU, 39.06 in YRI, and 32.22 in CHB, but was reduced to 11.58 in ASW. Before the release of CHM13 (Hoyt et al., 2022), chromosome 11 had the most completely sequenced centromere region, which had much rarer recombination events, all four cohorts showed an strong LD around the centromere (46,061,947-59,413,484 bp, totaling 1,035 unique grids) regardless of their ethnicities (Figure 4). , and 0.0022, respectively, and , and 0.094, respectively; the ratio for , and 94,05, for CEU, CHB, and YRI, respectively; the lowest ratio was found in ASW of 42.73. In addition, removing the HLA region of chromosome 6 or the centromere region of chromosome 11 would significantly reduce or in comparison with the randomly removal of other regions (Figure S5).

Model-based LD decay regression revealed LD composition

The real LD block size was not exact of m =250 or m =500, but an unknown parameter that should be inferred in computational intensive “LD decay” analysis (Zhang et al., 2019; Chang et al., 2015). We conducted the conventional LD decay for the 26 1KG cohorts (Figure 5A), and the time cost was 1,491.94 seconds for CHB. For each cohort, we took the area under the LD decay curve in the LD decay plot, and it quantified approximately the LD decay score for each cohort. The smallest score was 0.0421 for MSL and the largest was 0.0598 for PEL (Table 5). However, this estimation was not taken into account the real extent of LD, so it was not precise enough to reflect the LD decay score. For example, for admixture population, such as American cohorts, the extent of LD would be longer.

In contrast, we proposed a model-based method, as given in Eq 3, which could estimate LD decay score (regression coefficient b₁) and long-distance LD score (intercept b₀) jointly. Given the estimated 22 (Extended data 1; Table 4 for four representative cohorts and Supplementary R code), we regressed each autosomal against its correspondingly inversion of SNP number, and all yielded positive slopes (Pearson’s correlation ℛ > 0.80, Table 5; Figure 5B), an observation that was consistent with genome-wide spread of recombination hotspots. This linear relationship could consequently be considered the norm for a relative homogenous population as observed in most 1KG cohorts (Figure S6), while for the all 2,503 1KG samples ℛ =0.55 only (Table 5), indicating that the population structure and possible differentiated recombination hotspots across ethnicities disturbed the assumption underlying Eq 3 and smeared the linearity. We extracted and for the 26 1KG cohorts for further analysis. The rates of LD decay score, as indicated by , within the African cohorts (AFR) were significantly faster than other continents, consistent with previous observation that African population had relative shorter LD (Gabriel et al., 2002); while subgroups within the American continent (AMR) tended to have extended LD range due to their admixed genetic composition (Table 5 and Figure 5B). Notably, the correlation between and the approximated LD decay score was ℛ > 0.88. The estimated were highly correlated with .

Table 4.

Table 5.

A common feature was universally relative high LD of chromosome 6 and 11 in the 26 1KG cohorts (Figure S6). We quantified the impact of chromosome 6 and 11 by leave-one-chromosome-out test in CEU, CHB, YRI, and ASW for details (Figure 6A and 6B), and found that dropping chromosome 6 off could lift ℛ on average by 0.017, and chromosome 11 by 0.046. One possible explanation was that the centromere regions of chromosomes 6 and 11 have been assembled more completely than other chromosomes before the completion of CHM13 (Hoyt et al., 2022), whereas meiotic recombination tended to be reduced around the centromeres (Hinch et al., 2019). We estimated ℓ;_i after having knocked out the centromere region (46,061,947-59,413,484 bp, chr 11) in CEU, CHB, YRI, and ASW, and chromosome 11 then did not deviate much from their respective fitted lines (Figure 6C). A notable exceptional pattern was found in ASW, the chromosome 8 of which had even more deviation than chromosome 11 (ℛ was 0.83 and 0.87 with and without chromosome 8 in leave-one-chromosome out test) (Figure 6B). The deviation of chromosome 8 of ASW was consistent even more SNPs were added (Figure S7). We also provided high-resolution LD grids illustration for chromosome 8 (163,436 SNPs, totaling 214,185 grids) of the four representative cohorts for more detailed virtualization (Figure 6D). ASW had , but 0.00075, 0.00069 and 0.00043 for CEU, CHB, and YRI, respectively.

Discussion

In this study, we present a computationally efficient method to estimate mean LD of genomic grids of many SNP pairs. Our LD analysis framework is based on GRM, which has been embedded in variance component analysis for complex traits and genomic selection (Goddard, 2009; Visscher et al., 2014; Chen, 2014). The key connection from GRM to LD is bridged via the transformation between n × n matrix and m × m matrix, in particular here via Isserlis’s theorem under the fourth-order moment (Isserlis, 1918). With this connection, the computational cost for estimating the mean LD of m × m SNP pairs is reduced from 𝒪(nm²) to 𝒪(n²m), and the statistical properties of the proposed method are derived in theory and validated in 1KG datasets. In addition, as the genotype matrix X is of limited entries {0, 1, 2}, assuming missing genotypes are imputed first, using Mailman algorithm the computational cost of GRM can be further reduced to (Liberty and Zucker, 2009). The largest data tested so far for the proposed method has the sample size of 10,640 and of more than 5 million SNPs, it can complete genomic LD analysis in 77,508.00 seconds (Table 1). The weakness of the proposed method is obvious that the algorithm remains slow when the sample size is large or the grid resolution is increased. With the availability of such as UK Biobank data (Bycroft et al., 2018), the proposed method may not be adequate, and much advanced methods, such as randomizated implementation for the proposed methods, are needed.

We also applied the proposed method into 1KG and revealed certain characteristics of the human genomes. Firstly, we found the ubiquitously existence of extended LD, which was likely emerged because of population structure, even very slightly, and admixture history. We quantified the and in 1KG, and as indicated by we found the inter-chromosomal LD was nearly an order lower than intra-chromosomal LD; for admixed cohorts, the ratio was much higher, even very close to each other such as in all 1KG samples. Secondly, variable recombination rates shaped peak of local LD. For example, the HLA region showed high LD in European and East Asian cohorts, but relatively low LD in such as YRI, consistent with their much longer population history. Thirdly, it existed general linear correlation between ℓ;_i and the inversion of the SNP number, a long-anticipated result that is as predicted with genome-wide spread of recombination hotspots (Hinch et al., 2019). One outlier of this linear norm was chromosome 11, which had so far most completely genotyped centromere and consequently had more elevated LD compared with other autosomes. We anticipate that with the release of CHM13 the linear correlation should be much closer to unity (Hoyt et al., 2022). Of note, under the variance component analysis for complex traits, it is often a positive correlation between the length of a chromosome (as surrogated by the number of SNPs) and the proportion of heritability explained (Chen et al., 2014).

In contrast, throughout the study recurrent outstanding observations were found in ASW. For example, in ASW the ratio of was substantially dropped down compared with that of CEU, CHB, or YRI as illustrated in Figure 4. Furthermore, chromosome 8 in ASW fluctuated upwards most from the linear correlation (Figure 6), and even after various analyses, such as expanding SNP numbers. One possible explanation may lay under the complex demographic history of ASW, which can be investigated and tested in additional African American samples or possible existence for epistatic fitness (Ni et al., 2020).

Data availability

Public genetic datasets used in this study can be freely downloaded from the following URLs.

1000 Genomes Project: https://wwwdev.ebi.ac.uk/eva/?eva-study=PRJEB30460

CONVERGE: https://wwwdev.ebi.ac.uk/eva/?eva-study=PRJNA289433

Acknowledgements

GBC conceived and initiated the study. XH, TNZ, and YL conducted simulation and analyzed data. GBC, XH, TNZ, GAQ, and JZ developed the software. GBC wrote the first draft of the paper. All authors contributed to the writing, discussion of the paper, and validation of the results. This work was supported by National Natural Science Foundation of China (31771392), 110202101032(JY-09), and GZY-ZJ-KJ-23001. The funders played no role in designing, preparation, and submission of the paper.

Conflict of Interests

None.