Table 2 Results from ten-temperature melting analysis of hENT1 variants.

From: IMPROvER: the Integral Membrane Protein Stability Selector

Module

Variant

\(T_m\) (\(^{\circ }\)C)\(^{\mathrm{a}}\)

\(T_m\) error (\(\pm ^{\circ }\)C)\(^{\mathrm{b}}\)

\(\Delta T_m\) (\(^{\circ }\)C)

\(\Delta T_m\) error (\(\pm ^{\circ }\)C)\(^{\mathrm{c}}\)

Repeats (n)

Status

n/a

Wild-type

44.6

0.7

0.0

0.9

12

n/a

Deep-sequence

I380V

42.0

0.1

− 2.5

0.6

2

Destabilisng

Deep-sequence

M306T

41.3

0.1

0.4

1.8

2

Neutral

Deep-sequence

G225V

42.1

0.1

− 1.0

0.9

2

Destabilisng

Deep-sequence

S321T

44.9

1.7

− 2.5

0.6

2

Destabilisng

Model-based

R233L

37.1

1.6

− 3.8

0.6

2

Destabilisng

Model-based

S152L

42.7

1.3

− 1.9

1.4

5

Destabilisng

Model-based

E247M

45.8

0.6

1.2

0.8

4

Stabilising

Model-based

N30F

46.8

0.4

2.2

0.8

5

Stabilising

Model-based

L27E

45.5

0.7

0.9

0.9

5

Stabilising

Data-driven

A401L

46.3

0.7

1.7

1.0

2

Stabilising

Data-driven

T336A

43.9

0.7

− 0.7

0.9

5

Neutral

Data-driven

Q246A

47.6

1.1

0.2

0.9

5

Neutral

Data-driven

F153A

40.8

0.1

− 7.5

1.7

2

Destabilisng

Data-driven

G207A

43.6

0.7

− 3.3

0.6

2

Destabilisng

Data-driven

A88L

45.2

0.7

0.8

0.9

5

Neutral

Data-driven

K263A

47.2

1.1

2.6

1.3

5

Stabilising

  1. \(^{\mathrm{a}}\)Average \(T_m\) was calculated from individual \(T_m\) estimated for each individual repeat by fitting with a four-parameter dose-response curve (variable slope) by non-linear least-squares fitting in the python package scipy.stats.
  2. \(^{\mathrm{b}}\) For n > 1: standard error of the mean (SEM) shown.
  3. \(^{\mathrm{c}}\) For n > 1: error calculated and propagated as detailed in “Methods and materials” section.