Is locus structure really important?

[devrankrg]

[
    {
        "LocusId": "BEAN1",
        "LocusStructure": "(TAAAA)*",
        "VariantType": "RareRepeat",
        "ReferenceRegion": "16:66490398-66490453",
        "RepeatUnit": "TAAAA",
        "Gene": "BEAN1",
        "Inheritance": "AD",
        "GeneRegion": "intron",
        "GeneId": "ENSG00000166546",
        "DiscoveryMethod": "L",
        "DiscoveryYear": 2009,
        "PathogenicMotif": "TGGAA  (+TAGAA)",
        "Diseases": [
            {
                "Symbol": "SCA31",
                "Name": "Spinocerebellar ataxia 31",
                "Inheritance": "AD",
                "OMIM": "117210",
                "PathogenicMin": 500
            }
        ],
        "OfftargetRegions": [
            "10:94531487-94532013",
            "13:71654237-71654677",
            "17:46999846-47000288",
            "20:17081089-17081531",
            "21:23710971-23711386",
            "22:21395520-21395928",
            "2:44013201-44013638",
            "4:21714633-21715248",
            "5:116726213-116726658",
            "5:55105820-55106484",
            "7:151316374-151316835",
            "7:75175257-75175679",
            "9:110701505-110702081",
            "9:25543141-25543554",
            "X:6802782-6803290"
        ],
        "MainReferenceRegion": "16:66490398-66490453"
    }
]

I am doing research on the BEAN1 repeat. This repeat may have different repeat sequences in the same region, such as "TAGAA, TGGAA, TAAAA". The repeat sequence also affects the pathogenicity of the repeat. When I want to investigate pathogenic repeat, changing the locus structure with pathogenic repeat unit (TGGAA) has almost no effect on the result. Even when I write something of the same length but meaningless, like "HASAN", the result is still the same. What does regex used to find repeats do and how does it work?

Repeat's gnomad link = https://gnomad.broadinstitute.org/short-tandem-repeat/BEAN1?dataset=gnomad_r4

#CHROM	POS	ID	REF	ALT	QUAL	FILTER	INFO	FORMAT	HG002.GRCh38.2x250
chr16	66490398	.	A	<STR8>,<STR17>	.	PASS	END=66490453;REF=11;RL=55;RU=TGGAA;VARID=BEAN1;REPID=BEAN1	GT:SO:REPCN:REPCI:ADSP:ADFL:ADIR:LC	1/2:SPANNING/SPANNING:8/17:8-8/17-17:19/18:20/24:0/0:61.009132
chr16	66490398	.	A	<STR8>,<STR17>	.	PASS	END=66490453;REF=11;RL=55;RU=TAAAA;VARID=BEAN1;REPID=BEAN1	GT:SO:REPCN:REPCI:ADSP:ADFL:ADIR:LC	1/2:SPANNING/SPANNING:8/17:8-8/17-17:20/16:19/25:0/0:60.867580
chr16	66490398	.	A	<STR8>,<STR17>	.	PASS	END=66490453;REF=11;RL=55;RU=TRRRA;VARID=BEAN1;REPID=BEAN1	GT:SO:REPCN:REPCI:ADSP:ADFL:ADIR:LC	1/2:SPANNING/SPANNING:8/17:8-8/17-17:20/16:19/25:0/0:60.867580
chr16	66490398	.	A	<STR8>,<STR17>	.	PASS	END=66490453;REF=11;RL=55;RU=HASAN,VARID=BEAN1;REPID=BEAN1	GT:SO:REPCN:REPCI:ADSP:ADFL:ADIR:LC	1/2:SPANNING/SPANNING:8/17:8-8/17-17:19/16:19/24:0/0:60.726027

Although the RU field changed, the result remained almost unchanged
I tried;
"LocusStructure": "(TAAAA)"
"LocusStructure": "(TGGAA)"
"LocusStructure": "(TRRRA)"
"LocusStructure": "(HASAN)"