Ignoring the 2.79% Other and splitting the 1.74% Mongol-Tungus evenly between "Mongol" and "Tungus", this is what I got in my 23mofang to MDLP K23b model:
44.92% South_East_Asian
29.39% Tungus_Altaic
20.32% Austronesian
1.03% East_Siberian
0.35% Paleo_Siberian
0.92% South_Central_Asian
0.43% EHG
0.41% South_Indian
0.71% Caucasian
0.40% Ancestral_Altaic (ANE-like?)
96.0% East Asian, 0.4383 on my N-S cline. This is more southern than the simulated Han averages for any of the northern provinces, and somewhat more northern than me. The non-East Asian percentage is a little lower than my Shanxi simulation and around the same as my Gansu simulation.
Razib Khan Dec 2022 GNXP blog post:
The Hui Muslims, Two Pulses Of “Western” Ancestry 1,000 And 500 Years Ago, Mostly Male Mediated
https://academic.oup.com/mbe/article/38/9/3804/6281079
According to this paper, 39.3% of Ningxia Hui have West Eurasian Y-DNA haplogroups. The mtDNA percentage is lower than X-chromosome, which makes sense since X chromosome ancestry in women is inherited from both parents. From Razib: