Predict the PCA scores for a gt_pca, either for the original data or
projecting new data.
Arguments
- object
the
gt_pcaobject- new_data
a gen_tibble if scores are requested for a new dataset
- project_method
a string taking the value of either "simple", "OADP" (Online Augmentation, Decomposition, and Procrustes (OADP) projection), or "least_squares" (as done by SMARTPCA)
- lsq_pcs
a vector of length two with the values of the two principal components to use for the least square fitting. Only relevant if
project_method = 'least_squares'- block_size
number of loci read simultaneously (larger values will speed up computation, but require more memory)
- n_cores
number of cores
- as_matrix
logical, whether to return the result as a matrix (default) or a tibble.
- ...
no used
Value
a matrix of predictions (in line with predict using a prcomp
object) or a tibble, with samples as rows and components as columns. The
number of components depends on how many were estimated in the gt_pca
object.
References
Zhang et al (2020). Fast and robust ancestry prediction using principal component analysis 36(11): 3439–3446.
Examples
# Create a gen_tibble of lobster genotypes
bed_file <-
system.file("extdata", "lobster", "lobster.bed", package = "tidypopgen")
lobsters <- gen_tibble(bed_file,
backingfile = tempfile("lobsters"),
quiet = TRUE
)
# Remove monomorphic loci and impute
lobsters <- lobsters %>% select_loci_if(loci_maf(genotypes) > 0)
lobsters <- gt_impute_simple(lobsters, method = "mode")
# Subset into two datasets: one original and one to predict
original_lobsters <- lobsters[c(1:150), ]
new_lobsters <- lobsters[c(151:176), ]
# Create PCA object
pca <- gt_pca_partialSVD(original_lobsters)
# Predict
predict(pca, new_data = new_lobsters, project_method = "simple")
#> .PC1 .PC2 .PC3 .PC4 .PC5 .PC6
#> Vig11 0.84495324 -1.32057095 0.95199040 1.94839846 0.80850195 -2.48726987
#> Vig12 -0.07065694 -1.50166991 1.47821253 -0.65881868 -0.32228405 0.50084850
#> Vig13 0.58238791 0.39219100 0.99630978 -0.63288948 -1.20847141 1.38919001
#> Vig14 -1.99602145 -2.72731004 2.62251475 -1.98680500 -0.34075528 -0.12146277
#> Vig15 2.26512567 -0.95914412 0.02818837 -0.50138973 0.62787179 -0.64795356
#> Vig16 -1.61997544 0.15563731 -0.40919334 0.11173586 0.50622350 1.11854701
#> Vig17 -0.89287211 -3.18719723 -1.20601118 2.34870824 -2.33942914 -1.56549736
#> Vig18 1.05375670 -6.10652671 -0.82400057 -0.09187650 0.69699406 -1.40561403
#> Vig19 -1.17999551 -2.76944741 1.15010211 -1.20380960 0.24458605 -0.33721468
#> Vig20 -1.13094344 -2.51297661 1.29598913 -1.95585394 -1.40074692 -1.72114107
#> Vig21 1.45168737 -3.78272355 -0.65105851 1.25919966 -0.05981051 -0.72402871
#> Vig22 3.21701452 -4.81619496 1.41010615 -0.35000653 1.17661661 -2.28860783
#> Vig23 -1.90947818 -2.65470446 -0.37492036 -0.29023475 -2.52320183 -0.30582156
#> Vig24 1.28030055 -0.07325206 1.09767261 0.03358915 -1.70200452 0.14763356
#> Vig25 -2.52427515 0.39525228 0.57686195 -1.08051656 -0.36361029 0.71910092
#> Vig26 -0.75644140 -4.01534680 1.22994589 1.73545488 -1.08699849 0.77562836
#> Vig27 1.70410788 -0.05110208 1.16705289 1.26292192 -0.14261284 -0.59887368
#> Vig28 -0.21966527 -1.09228344 1.82206389 -0.11802565 -0.81543990 -1.93059542
#> Vig29 -1.80466347 -1.03162891 2.07763585 -0.30734296 1.41805841 -1.44704073
#> Vig30 0.69482871 -0.79526772 0.91429649 -0.73744265 0.63946822 -0.00832335
#> Vig31 0.28195945 -6.36122988 0.56220908 0.58385324 0.78660638 -2.00252223
#> Vig32 -0.93086395 -2.42914291 0.98211553 -0.10126105 0.17805143 0.96874494
#> Vig33 -0.37248494 -3.62456011 0.65785222 0.40916552 -0.56166707 0.26276509
#> Vig34 -0.30753874 -6.81517053 -1.18709724 -0.09927797 -1.72030191 -1.27847587
#> Vig35 -2.43398848 0.60635980 -1.25686251 0.32353264 -0.15047601 1.41621901
#> Vig36 1.46467362 -4.46678967 2.53264029 -1.12870891 -0.20734613 -1.62905258
#> .PC7 .PC8 .PC9 .PC10
#> Vig11 1.52777096 -1.51131063 0.74038292 0.01516207
#> Vig12 -0.12498898 -1.68443455 0.56254200 -0.48023226
#> Vig13 -0.02183661 1.28402899 -0.53105701 0.39991004
#> Vig14 -1.09481998 -1.78278611 0.72962139 0.74706789
#> Vig15 2.52704695 -0.81028103 1.56008567 1.21557644
#> Vig16 2.00582704 0.54789136 -0.95219207 0.20802815
#> Vig17 1.66238380 0.50856597 -0.03069758 1.13355924
#> Vig18 -0.36586879 -1.80176553 1.02610432 -1.85184381
#> Vig19 -0.31767420 -0.73445721 -0.22502040 -0.21155048
#> Vig20 2.10001332 -0.18007731 -0.76794917 0.02148115
#> Vig21 2.14159315 -2.00469523 -1.51150345 0.26520736
#> Vig22 0.35367486 -0.92968102 1.07611369 -2.00758176
#> Vig23 1.66242946 0.01520255 1.36135538 0.81743107
#> Vig24 1.11196915 0.56990781 0.16502708 1.12759456
#> Vig25 0.79085512 1.63079533 0.33625704 1.32880574
#> Vig26 -0.80559424 0.35422467 -0.02250144 -3.08140603
#> Vig27 1.93681287 -1.54261398 1.95076197 1.56666492
#> Vig28 0.70739380 -0.95388559 1.45470073 0.38913439
#> Vig29 0.12506807 0.88350450 0.31998914 0.27068806
#> Vig30 0.42978594 -0.75662056 0.38966104 0.22344619
#> Vig31 0.72805753 0.71855592 0.24921606 0.42971234
#> Vig32 -0.35527173 -1.01854014 -0.55387443 1.04468276
#> Vig33 0.02694817 -1.19006766 0.27706970 -2.34531402
#> Vig34 0.53566592 1.62395351 0.75212462 -0.61787774
#> Vig35 0.65649555 -1.25456935 -0.63035895 -0.19707490
#> Vig36 2.06423654 -1.37240744 1.98663932 -0.75062139
# Predict with OADP
predict(pca, new_data = new_lobsters, project_method = "OADP")
#> .PC1 .PC2 .PC3 .PC4 .PC5 .PC6
#> Vig11 0.89730930 -1.46636303 1.14351172 2.4359043 1.01915444 -3.1448299
#> Vig12 -0.07420107 -1.63420957 1.71061677 -0.7861686 -0.38700249 0.6028120
#> Vig13 0.62705559 0.44619604 1.25001301 -0.8350496 -1.61118140 1.8592399
#> Vig14 -2.08080051 -2.93066453 2.96913603 -2.3086783 -0.39806609 -0.1421679
#> Vig15 2.41022742 -1.06868561 0.03406264 -0.6314546 0.79751036 -0.8256000
#> Vig16 -1.72395889 0.17340689 -0.49408163 0.1405351 0.64206631 1.4230896
#> Vig17 -0.95308173 -3.57268268 -1.47493051 3.0038994 -3.01982148 -2.0276899
#> Vig18 1.11663168 -6.75697692 -0.98416390 -0.1140875 0.87244710 -1.7646337
#> Vig19 -1.24474568 -3.03780815 1.34981119 -1.4616112 0.29903569 -0.4133387
#> Vig20 -1.18708181 -2.73309619 1.49915896 -2.3335754 -1.68188082 -2.0713907
#> Vig21 1.53291181 -4.15843943 -0.76801709 1.5393596 -0.07365531 -0.8940289
#> Vig22 3.33481205 -5.12508629 1.56964673 -0.3984072 1.34543129 -2.6213354
#> Vig23 -2.01596603 -2.91707173 -0.44175403 -0.3542353 -3.10193978 -0.3769667
#> Vig24 1.37208979 -0.08264588 1.35614475 0.0434720 -2.22400661 0.1935946
#> Vig25 -2.64033034 0.42714612 0.65939444 -1.2701136 -0.42985734 0.8518972
#> Vig26 -0.79477312 -4.37544584 1.42836887 2.0811726 -1.31214200 0.9385453
#> Vig27 1.77091305 -0.05460700 1.30822356 1.4497249 -0.16450152 -0.6920186
#> Vig28 -0.23248841 -1.20534202 2.16272138 -0.1453498 -1.01184624 -2.4022659
#> Vig29 -1.91695225 -1.14571184 2.49432338 -0.3838393 1.78543395 -1.8273672
#> Vig30 0.73195749 -0.87012898 1.06793256 -0.8899979 0.77693514 -0.0101375
#> Vig31 0.29032128 -6.68971489 0.61286540 0.6477985 0.87589060 -2.2327502
#> Vig32 -0.97131000 -2.61395894 1.11403133 -0.1179058 0.20842652 1.1362299
#> Vig33 -0.39104264 -3.94297198 0.76113873 0.4882581 -0.67448688 0.3162787
#> Vig34 -0.32650580 -7.56719576 -1.42692375 -0.1242634 -2.17127575 -1.6185803
#> Vig35 -2.58340808 0.67242234 -1.50457785 0.4025691 -0.18872917 1.7814357
#> Vig36 1.53813454 -4.86368857 2.93793415 -1.3518221 -0.24996330 -1.9686106
#> .PC7 .PC8 .PC9 .PC10
#> Vig11 1.95234868 -1.9642099 0.98197318 0.02028417
#> Vig12 -0.15165556 -2.0700794 0.70196060 -0.60314481
#> Vig13 -0.02962272 1.7797042 -0.75539295 0.57518975
#> Vig14 -1.29022953 -2.1237486 0.88040369 0.90637989
#> Vig15 3.25554347 -1.0622741 2.08867165 1.64207875
#> Vig16 2.57978692 0.7168963 -1.27191013 0.28033492
#> Vig17 2.17912415 0.6794603 -0.04196384 1.56483356
#> Vig18 -0.46409685 -2.3233174 1.34945373 -2.45593814
#> Vig19 -0.39289152 -0.9213192 -0.28709492 -0.27186024
#> Vig20 2.54807945 -0.2213368 -0.95857439 0.02698959
#> Vig21 2.66956533 -2.5366278 -1.94721012 0.34427612
#> Vig22 0.40746993 -1.0810077 1.26508571 -2.37112079
#> Vig23 2.06837911 0.0191963 1.74966644 1.05853075
#> Vig24 1.47636775 0.7717487 0.22884625 1.57960635
#> Vig25 0.94380702 1.9688776 0.41161891 1.63618425
#> Vig26 -0.98311420 0.4381242 -0.02828174 -3.89951777
#> Vig27 2.25202988 -1.8113689 2.31756977 1.87049387
#> Vig28 0.88885861 -1.2172637 1.89109084 0.50987644
#> Vig29 0.15960429 1.1463640 0.42355155 0.36135013
#> Vig30 0.52799439 -0.9422649 0.49323663 0.28480387
#> Vig31 0.81550363 0.8107047 0.28360218 0.49078167
#> Vig32 -0.41956110 -1.2159306 -0.66979029 1.27023446
#> Vig33 0.03270156 -1.4628520 0.34585806 -2.94677444
#> Vig34 0.68551631 2.1140967 0.99945645 -0.82828812
#> Vig35 0.83430956 -1.6204983 -0.83024434 -0.26173005
#> Vig36 2.51565901 -1.6950287 2.49319130 -0.94843772
# Predict with least squares
predict(pca,
new_data = new_lobsters,
project_method = "least_squares", lsq_pcs = c(1, 2)
)
#> .PC1 .PC2
#> Vig11 0.8449532 -1.32057095
#> Vig12 -0.1088606 -1.79996190
#> Vig13 0.5823879 0.39219100
#> Vig14 -1.9960214 -2.72731004
#> Vig15 2.2651257 -0.95914412
#> Vig16 -1.6199754 0.15563731
#> Vig17 -1.0314321 -3.23991878
#> Vig18 1.0537567 -6.10652671
#> Vig19 -1.1799955 -2.76944741
#> Vig20 -1.1309434 -2.51297661
#> Vig21 1.4516874 -3.78272355
#> Vig22 3.2170145 -4.81619496
#> Vig23 -1.9094782 -2.65470446
#> Vig24 1.2803005 -0.07325206
#> Vig25 -2.5242751 0.39525228
#> Vig26 -0.7564414 -4.01534680
#> Vig27 1.7041079 -0.05110208
#> Vig28 -0.2233053 -1.09747960
#> Vig29 -1.8046635 -1.03162891
#> Vig30 0.6934134 -0.79727293
#> Vig31 0.2819594 -6.36122988
#> Vig32 -0.9308639 -2.42914291
#> Vig33 -0.3724849 -3.62456011
#> Vig34 -0.2919582 -6.89773149
#> Vig35 -2.4364457 0.61938077
#> Vig36 1.4646736 -4.46678967
# Return a tibble
predict(pca, new_data = new_lobsters, as_matrix = FALSE)
#> # A tibble: 26 × 11
#> id .PC1 .PC2 .PC3 .PC4 .PC5 .PC6 .PC7 .PC8 .PC9
#> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 Vig11 0.845 -1.32 0.952 1.95 0.809 -2.49 1.53 -1.51 0.740
#> 2 Vig12 -0.746 -1.19 1.34 -0.726 -0.481 0.306 -0.180 -1.88 0.511
#> 3 Vig13 0.582 0.392 0.996 -0.633 -1.21 1.39 -0.0218 1.28 -0.531
#> 4 Vig14 -2.00 -2.73 2.62 -1.99 -0.341 -0.121 -1.09 -1.78 0.730
#> 5 Vig15 2.27 -0.959 0.0282 -0.501 0.628 -0.648 2.53 -0.810 1.56
#> 6 Vig16 -1.62 0.156 -0.409 0.112 0.506 1.12 2.01 0.548 -0.952
#> 7 Vig17 -1.11 -3.27 -1.24 2.28 -2.37 -1.54 1.67 0.428 -0.176
#> 8 Vig18 1.05 -6.11 -0.824 -0.0919 0.697 -1.41 -0.366 -1.80 1.03
#> 9 Vig19 -1.18 -2.77 1.15 -1.20 0.245 -0.337 -0.318 -0.734 -0.225
#> 10 Vig20 -1.13 -2.51 1.30 -1.96 -1.40 -1.72 2.10 -0.180 -0.768
#> # ℹ 16 more rows
#> # ℹ 1 more variable: .PC10 <dbl>
# Adjust block.size
predict(pca, new_data = new_lobsters, block_size = 10)
#> .PC1 .PC2 .PC3 .PC4 .PC5 .PC6
#> Vig11 0.8449532 -1.32057095 0.95199040 1.94839846 0.80850195 -2.48726987
#> Vig12 -0.7461382 -1.18627774 1.34460570 -0.72604859 -0.48069122 0.30572175
#> Vig13 0.5823879 0.39219100 0.99630978 -0.63288948 -1.20847141 1.38919001
#> Vig14 -1.9960214 -2.72731004 2.62251475 -1.98680500 -0.34075528 -0.12146277
#> Vig15 2.2651257 -0.95914412 0.02818837 -0.50138973 0.62787179 -0.64795356
#> Vig16 -1.6199754 0.15563731 -0.40919334 0.11173586 0.50622350 1.11854701
#> Vig17 -1.1110388 -3.27020882 -1.23906625 2.28063954 -2.37335730 -1.54383228
#> Vig18 1.0537567 -6.10652671 -0.82400057 -0.09187650 0.69699406 -1.40561403
#> Vig19 -1.1799955 -2.76944741 1.15010211 -1.20380960 0.24458605 -0.33721468
#> Vig20 -1.1309434 -2.51297661 1.29598913 -1.95585394 -1.40074692 -1.72114107
#> Vig21 1.4516874 -3.78272355 -0.65105851 1.25919966 -0.05981051 -0.72402871
#> Vig22 3.2170145 -4.81619496 1.41010615 -0.35000653 1.17661661 -2.28860783
#> Vig23 -1.9094782 -2.65470446 -0.37492036 -0.29023475 -2.52320183 -0.30582156
#> Vig24 1.2803005 -0.07325206 1.09767261 0.03358915 -1.70200452 0.14763356
#> Vig25 -2.5242751 0.39525228 0.57686195 -1.08051656 -0.36361029 0.71910092
#> Vig26 -0.7564414 -4.01534680 1.22994589 1.73545488 -1.08699849 0.77562836
#> Vig27 1.7041079 -0.05110208 1.16705289 1.26292192 -0.14261284 -0.59887368
#> Vig28 -0.1765855 -1.03078650 1.97855185 -0.22492693 -0.96831542 -2.02312805
#> Vig29 -1.8046635 -1.03162891 2.07763585 -0.30734296 1.41805841 -1.44704073
#> Vig30 0.7479859 -0.71995438 1.08461798 -0.80963486 0.63013991 -0.05207368
#> Vig31 0.2819594 -6.36122988 0.56220908 0.58385324 0.78660638 -2.00252223
#> Vig32 -0.9308639 -2.42914291 0.98211553 -0.10126105 0.17805143 0.96874494
#> Vig33 -0.3724849 -3.62456011 0.65785222 0.40916552 -0.56166707 0.26276509
#> Vig34 -0.3262743 -6.71589081 -1.14051137 -0.12775555 -1.68695822 -1.41827318
#> Vig35 -2.4527241 0.70563952 -1.21027664 0.29505506 -0.11713232 1.27642171
#> Vig36 1.4646736 -4.46678967 2.53264029 -1.12870891 -0.20734613 -1.62905258
#> .PC7 .PC8 .PC9 .PC10
#> Vig11 1.52777096 -1.51131063 0.74038292 0.01516207
#> Vig12 -0.17965094 -1.88387628 0.51083352 -0.33320105
#> Vig13 -0.02183661 1.28402899 -0.53105701 0.39991004
#> Vig14 -1.09481998 -1.78278611 0.72962139 0.74706789
#> Vig15 2.52704695 -0.81028103 1.56008567 1.21557644
#> Vig16 2.00582704 0.54789136 -0.95219207 0.20802815
#> Vig17 1.67054359 0.42798506 -0.17562231 1.20681868
#> Vig18 -0.36586879 -1.80176553 1.02610432 -1.85184381
#> Vig19 -0.31767420 -0.73445721 -0.22502040 -0.21155048
#> Vig20 2.10001332 -0.18007731 -0.76794917 0.02148115
#> Vig21 2.14159315 -2.00469523 -1.51150345 0.26520736
#> Vig22 0.35367486 -0.92968102 1.07611369 -2.00758176
#> Vig23 1.66242946 0.01520255 1.36135538 0.81743107
#> Vig24 1.11196915 0.56990781 0.16502708 1.12759456
#> Vig25 0.79085512 1.63079533 0.33625704 1.32880574
#> Vig26 -0.80559424 0.35422467 -0.02250144 -3.08140603
#> Vig27 1.93681287 -1.54261398 1.95076197 1.56666492
#> Vig28 0.56401377 -0.97511516 1.32664358 0.44977954
#> Vig29 0.12506807 0.88350450 0.31998914 0.27068806
#> Vig30 0.41996665 -0.90250715 0.30991115 0.25806489
#> Vig31 0.72805753 0.71855592 0.24921606 0.42971234
#> Vig32 -0.35527173 -1.01854014 -0.55387443 1.04468276
#> Vig33 0.02694817 -1.19006766 0.27706970 -2.34531402
#> Vig34 0.51631535 1.67658793 0.68897116 -0.66022471
#> Vig35 0.63714498 -1.20193493 -0.69351240 -0.23942188
#> Vig36 2.06423654 -1.37240744 1.98663932 -0.75062139