Comparison of HapMap and 1000 Genomes Reference Panels in a Large-Scale Genome-Wide Association Study

Paul S de Vries; Maria Sabater-Lleal; Daniel I Chasman; Stella Trompet; Tarunveer S Ahluwalia; Alexander Teumer; Marcus E Kleber; Ming-Huei Chen; Jie Jin Wang; John R Attia; Riccardo E Marioni; Maristella Steri; Lu-Chen Weng; Rene Pool; Vera Grossmann; Jennifer A Brody; Cristina Venturini; Toshiko Tanaka; Lynda M Rose; Christopher Oldmeadow; Johanna Mazur; Saonli Basu; Mattias Frånberg; Qiong Yang; Symen Ligthart; Jouke J Hottenga; Ann Rumley; Antonella Mulas; Anton J M de Craen; Anne Grotevendt; Kent D Taylor; Graciela E Delgado; Annette Kifley; Lorna M Lopez; Tina L Berentzen; Massimo Mangino; Stefania Bandinelli; Alanna C Morrison; Anders Hamsten; Geoffrey Tofler; Moniek P M de Maat; Harmen H M Draisma; Gordon D Lowe; Magdalena Zoledziewska; Naveed Sattar; Karl J Lackner; Uwe Völker; Barbara McKnight; Jie Huang; Elizabeth G Holliday; Mark A McEvoy; John M Starr; Pirro G Hysi; Dena G Hernandez; Weihua Guan; Fernando Rivadeneira; Wendy L McArdle; P Eline Slagboom; Tanja Zeller; Bruce M Psaty; André G Uitterlinden; Eco J C de Geus; David J Stott; Harald Binder; Albert Hofman; Oscar H Franco; Jerome I Rotter; Luigi Ferrucci; Tim D Spector; Ian J Deary; Winfried März; Andreas Greinacher; Philipp S Wild; Francesco Cucca; Dorret I Boomsma; Hugh Watkins; Weihong Tang; Paul M Ridker; Jan W Jukema; Rodney J Scott; Paul Mitchell; Torben Hansen; Christopher J O'Donnell; Nicholas L Smith; David P Strachan; Abbas Dehghan

doi:10.1371/journal.pone.0167742

Comparison of HapMap and 1000 Genomes Reference Panels in a Large-Scale Genome-Wide Association Study

Paul S de Vries, Maria Sabater-Lleal, Daniel I Chasman, Stella Trompet, Tarunveer S Ahluwalia, Alexander Teumer, Marcus E Kleber, Ming-Huei Chen, Jie Jin Wang, John R Attia, Riccardo E Marioni, Maristella Steri, Lu-Chen Weng, Rene Pool, Vera Grossmann, Jennifer A Brody, Cristina Venturini, Toshiko Tanaka, Lynda M Rose, Christopher OldmeadowJohanna Mazur, Saonli Basu, Mattias Frånberg, Qiong Yang, Symen Ligthart, Jouke J Hottenga, Ann Rumley, Antonella Mulas, Anton J M de Craen, Anne Grotevendt, Kent D Taylor, Graciela E Delgado, Annette Kifley, Lorna M Lopez, Tina L Berentzen, Massimo Mangino, Stefania Bandinelli, Alanna C Morrison, Anders Hamsten, Geoffrey Tofler, Moniek P M de Maat, Harmen H M Draisma, Gordon D Lowe, Magdalena Zoledziewska, Naveed Sattar, Karl J Lackner, Uwe Völker, Barbara McKnight, Jie Huang, Elizabeth G Holliday, Mark A McEvoy, John M Starr, Pirro G Hysi, Dena G Hernandez, Weihua Guan, Fernando Rivadeneira, Wendy L McArdle, P Eline Slagboom, Tanja Zeller, Bruce M Psaty, André G Uitterlinden, Eco J C de Geus, David J Stott, Harald Binder, Albert Hofman, Oscar H Franco, Jerome I Rotter, Luigi Ferrucci, Tim D Spector, Ian J Deary, Winfried März, Andreas Greinacher, Philipp S Wild, Francesco Cucca, Dorret I Boomsma, Hugh Watkins, Weihong Tang, Paul M Ridker, Jan W Jukema, Rodney J Scott, Paul Mitchell, Torben Hansen, Christopher J O'Donnell, Nicholas L Smith, David P Strachan, Abbas Dehghan

16 Citations (Scopus)

49 Downloads (Pure)

Abstract

An increasing number of genome-wide association (GWA) studies are now using the higher resolution 1000 Genomes Project reference panel (1000G) for imputation, with the expectation that 1000G imputation will lead to the discovery of additional associated loci when compared to HapMap imputation. In order to assess the improvement of 1000G over HapMap imputation in identifying associated loci, we compared the results of GWA studies of circulating fibrinogen based on the two reference panels. Using both HapMap and 1000G imputation we performed a meta-analysis of 22 studies comprising the same 91,953 individuals. We identified six additional signals using 1000G imputation, while 29 loci were associated using both HapMap and 1000G imputation. One locus identified using HapMap imputation was not significant using 1000G imputation. The genome-wide significance threshold of 5×10^-8 is based on the number of independent statistical tests using HapMap imputation, and 1000G imputation may lead to further independent tests that should be corrected for. When using a stricter Bonferroni correction for the 1000G GWA study (P-value < 2.5×10^-8), the number of loci significant only using HapMap imputation increased to 4 while the number of loci significant only using 1000G decreased to 5. In conclusion, 1000G imputation enabled the identification of 20% more loci than HapMap imputation, although the advantage of 1000G imputation became less clear when a stricter Bonferroni correction was used. More generally, our results provide insights that are applicable to the implementation of other dense reference panels that are under development.

Original language	English
Article number	e0167742
Journal	PLOS ONE
Volume	12
Issue number	1
Number of pages	22
ISSN	1932-6203
DOIs	https://doi.org/10.1371/journal.pone.0167742
Publication status	Published - Jan 2017

Keywords

Journal Article

Access to Document

10.1371/journal.pone.0167742Licence: CC0

fileFinal published version, 1.41 MBLicence: CC0

Cite this

de Vries, P. S., Sabater-Lleal, M., Chasman, D. I., Trompet, S., Ahluwalia, T. S., Teumer, A., Kleber, M. E., Chen, M-H., Wang, J. J., Attia, J. R., Marioni, R. E., Steri, M., Weng, L-C., Pool, R., Grossmann, V., Brody, J. A., Venturini, C., Tanaka, T., Rose, L. M., ... Dehghan, A. (2017). Comparison of HapMap and 1000 Genomes Reference Panels in a Large-Scale Genome-Wide Association Study. PLOS ONE, 12(1), [e0167742]. https://doi.org/10.1371/journal.pone.0167742

de Vries, PS, Sabater-Lleal, M, Chasman, DI, Trompet, S, Ahluwalia, TS, Teumer, A, Kleber, ME, Chen, M-H, Wang, JJ, Attia, JR, Marioni, RE, Steri, M, Weng, L-C, Pool, R, Grossmann, V, Brody, JA, Venturini, C, Tanaka, T, Rose, LM, Oldmeadow, C, Mazur, J, Basu, S, Frånberg, M, Yang, Q, Ligthart, S, Hottenga, JJ, Rumley, A, Mulas, A, de Craen, AJM, Grotevendt, A, Taylor, KD, Delgado, GE, Kifley, A, Lopez, LM, Berentzen, TL, Mangino, M, Bandinelli, S, Morrison, AC, Hamsten, A, Tofler, G, de Maat, MPM, Draisma, HHM, Lowe, GD, Zoledziewska, M, Sattar, N, Lackner, KJ, Völker, U, McKnight, B, Huang, J, Holliday, EG, McEvoy, MA, Starr, JM, Hysi, PG, Hernandez, DG, Guan, W, Rivadeneira, F, McArdle, WL, Slagboom, PE, Zeller, T, Psaty, BM, Uitterlinden, AG, de Geus, EJC, Stott, DJ, Binder, H, Hofman, A, Franco, OH, Rotter, JI, Ferrucci, L, Spector, TD, Deary, IJ, März, W, Greinacher, A, Wild, PS, Cucca, F, Boomsma, DI, Watkins, H, Tang, W, Ridker, PM, Jukema, JW, Scott, RJ, Mitchell, P, Hansen, T, O'Donnell, CJ, Smith, NL, Strachan, DP & Dehghan, A 2017, 'Comparison of HapMap and 1000 Genomes Reference Panels in a Large-Scale Genome-Wide Association Study', PLOS ONE, vol. 12, no. 1, e0167742. https://doi.org/10.1371/journal.pone.0167742

@article{423fc2ab748542f986bbcfdd34c7db0e,

title = "Comparison of HapMap and 1000 Genomes Reference Panels in a Large-Scale Genome-Wide Association Study",

abstract = "An increasing number of genome-wide association (GWA) studies are now using the higher resolution 1000 Genomes Project reference panel (1000G) for imputation, with the expectation that 1000G imputation will lead to the discovery of additional associated loci when compared to HapMap imputation. In order to assess the improvement of 1000G over HapMap imputation in identifying associated loci, we compared the results of GWA studies of circulating fibrinogen based on the two reference panels. Using both HapMap and 1000G imputation we performed a meta-analysis of 22 studies comprising the same 91,953 individuals. We identified six additional signals using 1000G imputation, while 29 loci were associated using both HapMap and 1000G imputation. One locus identified using HapMap imputation was not significant using 1000G imputation. The genome-wide significance threshold of 5×10-8 is based on the number of independent statistical tests using HapMap imputation, and 1000G imputation may lead to further independent tests that should be corrected for. When using a stricter Bonferroni correction for the 1000G GWA study (P-value < 2.5×10-8), the number of loci significant only using HapMap imputation increased to 4 while the number of loci significant only using 1000G decreased to 5. In conclusion, 1000G imputation enabled the identification of 20% more loci than HapMap imputation, although the advantage of 1000G imputation became less clear when a stricter Bonferroni correction was used. More generally, our results provide insights that are applicable to the implementation of other dense reference panels that are under development.",

keywords = "Journal Article",

author = "{de Vries}, {Paul S} and Maria Sabater-Lleal and Chasman, {Daniel I} and Stella Trompet and Ahluwalia, {Tarunveer S} and Alexander Teumer and Kleber, {Marcus E} and Ming-Huei Chen and Wang, {Jie Jin} and Attia, {John R} and Marioni, {Riccardo E} and Maristella Steri and Lu-Chen Weng and Rene Pool and Vera Grossmann and Brody, {Jennifer A} and Cristina Venturini and Toshiko Tanaka and Rose, {Lynda M} and Christopher Oldmeadow and Johanna Mazur and Saonli Basu and Mattias Fr{\aa}nberg and Qiong Yang and Symen Ligthart and Hottenga, {Jouke J} and Ann Rumley and Antonella Mulas and {de Craen}, {Anton J M} and Anne Grotevendt and Taylor, {Kent D} and Delgado, {Graciela E} and Annette Kifley and Lopez, {Lorna M} and Berentzen, {Tina L} and Massimo Mangino and Stefania Bandinelli and Morrison, {Alanna C} and Anders Hamsten and Geoffrey Tofler and {de Maat}, {Moniek P M} and Draisma, {Harmen H M} and Lowe, {Gordon D} and Magdalena Zoledziewska and Naveed Sattar and Lackner, {Karl J} and Uwe V{\"o}lker and Barbara McKnight and Jie Huang and Holliday, {Elizabeth G} and McEvoy, {Mark A} and Starr, {John M} and Hysi, {Pirro G} and Hernandez, {Dena G} and Weihua Guan and Fernando Rivadeneira and McArdle, {Wendy L} and Slagboom, {P Eline} and Tanja Zeller and Psaty, {Bruce M} and Uitterlinden, {Andr{\'e} G} and {de Geus}, {Eco J C} and Stott, {David J} and Harald Binder and Albert Hofman and Franco, {Oscar H} and Rotter, {Jerome I} and Luigi Ferrucci and Spector, {Tim D} and Deary, {Ian J} and Winfried M{\"a}rz and Andreas Greinacher and Wild, {Philipp S} and Francesco Cucca and Boomsma, {Dorret I} and Hugh Watkins and Weihong Tang and Ridker, {Paul M} and Jukema, {Jan W} and Scott, {Rodney J} and Paul Mitchell and Torben Hansen and O'Donnell, {Christopher J} and Smith, {Nicholas L} and Strachan, {David P} and Abbas Dehghan",

year = "2017",

month = jan,

doi = "10.1371/journal.pone.0167742",

language = "English",

volume = "12",

journal = "PLoS Computational Biology",

issn = "1932-6203",

publisher = "Public Library of Science",

number = "1",

}

TY - JOUR

T1 - Comparison of HapMap and 1000 Genomes Reference Panels in a Large-Scale Genome-Wide Association Study

AU - de Vries, Paul S

AU - Sabater-Lleal, Maria

AU - Chasman, Daniel I

AU - Trompet, Stella

AU - Ahluwalia, Tarunveer S

AU - Teumer, Alexander

AU - Kleber, Marcus E

AU - Chen, Ming-Huei

AU - Wang, Jie Jin

AU - Attia, John R

AU - Marioni, Riccardo E

AU - Steri, Maristella

AU - Weng, Lu-Chen

AU - Pool, Rene

AU - Grossmann, Vera

AU - Brody, Jennifer A

AU - Venturini, Cristina

AU - Tanaka, Toshiko

AU - Rose, Lynda M

AU - Oldmeadow, Christopher

AU - Mazur, Johanna

AU - Basu, Saonli

AU - Frånberg, Mattias

AU - Yang, Qiong

AU - Ligthart, Symen

AU - Hottenga, Jouke J

AU - Rumley, Ann

AU - Mulas, Antonella

AU - de Craen, Anton J M

AU - Grotevendt, Anne

AU - Taylor, Kent D

AU - Delgado, Graciela E

AU - Kifley, Annette

AU - Lopez, Lorna M

AU - Berentzen, Tina L

AU - Mangino, Massimo

AU - Bandinelli, Stefania

AU - Morrison, Alanna C

AU - Hamsten, Anders

AU - Tofler, Geoffrey

AU - de Maat, Moniek P M

AU - Draisma, Harmen H M

AU - Lowe, Gordon D

AU - Zoledziewska, Magdalena

AU - Sattar, Naveed

AU - Lackner, Karl J

AU - Völker, Uwe

AU - McKnight, Barbara

AU - Huang, Jie

AU - Holliday, Elizabeth G

AU - McEvoy, Mark A

AU - Starr, John M

AU - Hysi, Pirro G

AU - Hernandez, Dena G

AU - Guan, Weihua

AU - Rivadeneira, Fernando

AU - McArdle, Wendy L

AU - Slagboom, P Eline

AU - Zeller, Tanja

AU - Psaty, Bruce M

AU - Uitterlinden, André G

AU - de Geus, Eco J C

AU - Stott, David J

AU - Binder, Harald

AU - Hofman, Albert

AU - Franco, Oscar H

AU - Rotter, Jerome I

AU - Ferrucci, Luigi

AU - Spector, Tim D

AU - Deary, Ian J

AU - März, Winfried

AU - Greinacher, Andreas

AU - Wild, Philipp S

AU - Cucca, Francesco

AU - Boomsma, Dorret I

AU - Watkins, Hugh

AU - Tang, Weihong

AU - Ridker, Paul M

AU - Jukema, Jan W

AU - Scott, Rodney J

AU - Mitchell, Paul

AU - Hansen, Torben

AU - O'Donnell, Christopher J

AU - Smith, Nicholas L

AU - Strachan, David P

AU - Dehghan, Abbas

PY - 2017/1

Y1 - 2017/1

N2 - An increasing number of genome-wide association (GWA) studies are now using the higher resolution 1000 Genomes Project reference panel (1000G) for imputation, with the expectation that 1000G imputation will lead to the discovery of additional associated loci when compared to HapMap imputation. In order to assess the improvement of 1000G over HapMap imputation in identifying associated loci, we compared the results of GWA studies of circulating fibrinogen based on the two reference panels. Using both HapMap and 1000G imputation we performed a meta-analysis of 22 studies comprising the same 91,953 individuals. We identified six additional signals using 1000G imputation, while 29 loci were associated using both HapMap and 1000G imputation. One locus identified using HapMap imputation was not significant using 1000G imputation. The genome-wide significance threshold of 5×10-8 is based on the number of independent statistical tests using HapMap imputation, and 1000G imputation may lead to further independent tests that should be corrected for. When using a stricter Bonferroni correction for the 1000G GWA study (P-value < 2.5×10-8), the number of loci significant only using HapMap imputation increased to 4 while the number of loci significant only using 1000G decreased to 5. In conclusion, 1000G imputation enabled the identification of 20% more loci than HapMap imputation, although the advantage of 1000G imputation became less clear when a stricter Bonferroni correction was used. More generally, our results provide insights that are applicable to the implementation of other dense reference panels that are under development.

AB - An increasing number of genome-wide association (GWA) studies are now using the higher resolution 1000 Genomes Project reference panel (1000G) for imputation, with the expectation that 1000G imputation will lead to the discovery of additional associated loci when compared to HapMap imputation. In order to assess the improvement of 1000G over HapMap imputation in identifying associated loci, we compared the results of GWA studies of circulating fibrinogen based on the two reference panels. Using both HapMap and 1000G imputation we performed a meta-analysis of 22 studies comprising the same 91,953 individuals. We identified six additional signals using 1000G imputation, while 29 loci were associated using both HapMap and 1000G imputation. One locus identified using HapMap imputation was not significant using 1000G imputation. The genome-wide significance threshold of 5×10-8 is based on the number of independent statistical tests using HapMap imputation, and 1000G imputation may lead to further independent tests that should be corrected for. When using a stricter Bonferroni correction for the 1000G GWA study (P-value < 2.5×10-8), the number of loci significant only using HapMap imputation increased to 4 while the number of loci significant only using 1000G decreased to 5. In conclusion, 1000G imputation enabled the identification of 20% more loci than HapMap imputation, although the advantage of 1000G imputation became less clear when a stricter Bonferroni correction was used. More generally, our results provide insights that are applicable to the implementation of other dense reference panels that are under development.

KW - Journal Article

U2 - 10.1371/journal.pone.0167742

DO - 10.1371/journal.pone.0167742

M3 - Journal article

C2 - 28107422

SN - 1932-6203

VL - 12

JO - PLoS Computational Biology

JF - PLoS Computational Biology

IS - 1

M1 - e0167742

ER -

Comparison of HapMap and 1000 Genomes Reference Panels in a Large-Scale Genome-Wide Association Study

Abstract

Keywords

Access to Document

Fingerprint

Cite this