I can't find any design issue in your post.
Your question is not design issue but simple usage of EDA tool.  
Please post to
"The Designer's Guide Community Forum ≫ Simulators ≫ RF Simulators".
Or
Agilent Home >Products & Services > Test & Measurement >  
Agilent EEsof EDA > EEsof Knowledge Center > Discussion Forums  
Agilent EEsof Forums > ADS > Analog Simulation
https://edasupportweb.soco.agilent.com/cgi-bin/yabb/YaBB.pl


This is packaged BJT model.
AT41486 have 4 pins(c, b, e, e).
You open one emitter pin.
Try connect two emitter pins.

And why don't you set C1 as 10pF//100pF.

I've found your mistake.
You've set wrong port number.
You shouls set num=1 for input port.
And set num=2 for output port.

You are using old version of AT41486("pb_hp_AT41486_19920721"). Maybe this old version is not proper since it gives 2.55dB as NFmin even though I invoke ideal measurement condition.
I don't know why Agilent keeps to bandle this old version. You should contact with Agilent EDA support.

As GummelPoon Model, there are following three versions.
 #uselib "RfTransistorLibrary" , "pb_hp_AT41486_19920721"
 #uselib "RfTransistorLibrary" , "pb_hp_AT41486_19921101"
 #uselib "RfTransistorLibrary" , "pb_hp_AT41486_19931202"

As S-parameter models, there are following two models.
 "sp_hp_AT-41486_1_19920721"
 "sp_hp_AT-41486_1_19921201"

If you use S-parameter models, you get completely 1.7dB as NFmin.

And if you use "pb_hp_AT41486_19921101" or "pb_hp_AT41486_19931202", you get almost 1.7dB as NFmin.