Correct publication references, input guidance and output units

README.md

@@ -9,7 +9,7 @@ tags:
 
 # mccartneyldlcholesterol
 
-Whole-blood DNAm LASSO score for LDL with remnant cholesterol, trained in Generation Scotland on an age-, sex-, and ancestry-adjusted phenotype residual and evaluated out of sample in LBC1936.
+Whole-blood DNAm LASSO score for LDL with remnant cholesterol, trained in Generation Scotland on an age-, sex-, and ancestry-adjusted phenotype residual and evaluated out of sample in LBC1936. The current pyaging implementation applies a sigmoid to the linear score, so the returned value is bounded and unitless. It is neither the original continuous phenotype residual nor a calibrated probability; the paper does not specify this sigmoid.
 
 Model weights retain the original authors' terms; the pyaging software license does not relicense them.
 

config.json

@@ -10,7 +10,7 @@
  "last_author": "Riccardo E. Marioni",
  "model_type": "LASSO regression",
  "n_features": 233,
- "notes": "Whole-blood DNAm LASSO score for LDL with remnant cholesterol, trained in Generation Scotland on an age-, sex-, and ancestry-adjusted phenotype residual and evaluated out of sample in LBC1936.",
+ "notes": "Whole-blood DNAm LASSO score for LDL with remnant cholesterol, trained in Generation Scotland on an age-, sex-, and ancestry-adjusted phenotype residual and evaluated out of sample in LBC1936. The current pyaging implementation applies a sigmoid to the linear score, so the returned value is bounded and unitless. It is neither the original continuous phenotype residual nor a calibrated probability; the paper does not specify this sigmoid.",
  "platform": [
   "Illumina EPIC"
  ],

mccartneyldlcholesterol.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:1e61804705d5049275fcf38495a6509981b341c1f4b3f11432949789d14d0ceb
-size 12389
+oid sha256:f58a25c59b0251cbbe88ff90e88606ff4c93020487f2665812eb3427d01efb3c
+size 12581