Table 5 Prediction error of the RF, ULSTM, BLSTM and CNN-LSTM models, as measured by RMSD in Å units add std on the tables. The intracellular loops (ICL) of the APO, FA, PIA simulations for 2rh1 and 3p0g states.

From: GPCR molecular dynamics forecasting using recurrent neural networks

Region

MD-RMSD

RF

ULSTM

BLSTM

CNN-LSTM

APO

2rh1

ICL1\(^{**+}\)

1.6690

0.2769\(^{**-}\) \({\pm }\) 0.0122

0.1669 \({\pm }\) 0.0085

0.1727 \({\pm }\) 0.0068

0.1699\({\pm }\) 0.0026

ICL2

1.9260

0.3104\(^{**-}\) \({\pm }\) 0.0135

0.1926 \({\pm }\) 0.0091

0.1980 \({\pm }\) 0.0069

0.1978 \({\pm }\) 0.0025

APO

3p0g

ICL1\(^{**+}\)

1.7470

0.1869\(^{*-}\) \({\pm }\) 0.0038

0.1747 \({\pm }\) 0.0057

0.1777 \({\pm }\) 0.0055

0.1717 \({\pm }\) 0.0017

ICL2

2.0260

0.2183\(^{*-}\) \({\pm }\) 0.0009

0.2026 \({\pm }\) 0.0057

0.2062 \({\pm }\) 0.0057

0.2038 \({\pm }\) 0.0022

FA

2rh1

ICL1\(^{**+}\)

1.7960

0.1945\(^{**-}\) \({\pm }\) 0.0024

0.1713 \({\pm }\) 0.0039

0.1725 \({\pm }\) 0.0038

0.1732 \({\pm }\) 0.0046

ICL2

1.3420

0.2466\(^{**-}\) \({\pm }\) 0.0033

0.1934 \({\pm }\) 0.0042

0.1978 \({\pm }\) 0.0043

0.1979 \({\pm }\) 0.0042

FA

3p0g

ICL1\(^{**+}\)

1.7510

0.1828\(^{*-}\) \({\pm }\) 0.0031

0.1751 \({\pm }\) 0.0050

0.1739 \({\pm }\) 0.0038

0.1704 \({\pm }\) 0.0034

ICL2

2.0170

0.2269\(^{**-}\) \({\pm }\) 0.0019

0.2017 \({\pm }\) 0.0065

0.1988 \({\pm }\) 0.0050

0.1964 \({\pm }\) 0.0032

PIA

2rh1

ICL1\(^{**+}\)

1.7770

0.2270\(^{**-}\) \({\pm }\) 0.0085

0.1777 \({\pm }\) 0.0022

0.1766 \({\pm }\) 0.0045

0.1703 \({\pm }\) 0.0102

ICL2

2.0020

0.2590\(^{**-}\) \({\pm }\) 0.0080

0.2002 \({\pm }\) 0.0030

0.2027 \({\pm }\) 0.0047

0.1996 \({\pm }\) 0.0087

PIA

3p0g

ICL1\(^{**+}\)

1.7460

0.1896\(^{*-}\) \({\pm }\) 0.0053

0.1746 \({\pm }\) 0.0078

0.1795 \({\pm }\) 0.0062

0.1766 \({\pm }\) 0.0020

ICL2

1.9670

0.2520\(^{**-}\) \({\pm }\) 0.0040

0.1967 \({\pm }\) 0.0074

0.2049 \({\pm }\) 0.0068

0.2013 \({\pm }\) 0.0023

Mean

ICL1\(^{**+}\)

1.7477

0.2096\(^{**-}\) \({\pm }\) 0.0059

0.1747 \({\pm }\) 0.0049

0.1755 \({\pm }\) 0.0051

0.1720 \({\pm }\) 0.0041

ICL2

1.8800

0.2522\(^{**-}\) \({\pm }\) 0.0053

0.1989 \({\pm }\) 0.0054

0.2014 \({\pm }\) 0.0056

0.1995 \({\pm }\) 0.0038

  1. Statistics of significantly worse\(^{(-)}\) and significantly better\(^{(+)}\) differences with p \(< 0.05^{*}\) and p\(< 0.01^{**}\). The symbols in the first column indicate significantly better differences between ICL regions, while, in the model results columns, they indicate significantly worse differences between models. The Mean row at the bottom of the table indicates mean errors across regions. Bold values show the minimum error accross regions and models.
Back to article page