If just one calibration point is available, the calibration can be done in one of two ways. An offset error can be removed from the ADC output values to obtain a rough compensation. Both the ADC value at T1 (room temperature) and at T2 can be used to find this offset error. Note that this kind of calibration/compensation could result in larger errors in the measurements. Let us for example assume that the ADC output has a negative error in the low end of the temperature scale, and that the error turns positive above a given temperature, around 50°C. If the calibration point is at 25°C and the temperatures of interest are above 50°C, compensation using only offset removal would result in larger measuring errors for the temperatures of interest.
The second type of one-point calibration is to assume that the output from the ADC is approximately linear and that the error is primarily a gain error. The calibration point is used to calculate the gain factor, when assuming zero offset. This kind of calibration and compensation suffers from the same problems as offset calibration.
The advantage of one-point calibration compared to two-point calibration is the fact that only one accurate calibration temperature is needed. Also, the compensation itself is simpler because the formula applied on the output values is less complicated.