Gene Array Subjects

SUBJECTS ARRANGED ACCORDING TO GROUP ASSIGNMENT

* Global Impairment Score is an index of overall performance on the NNTC neuropsychological test battery. A score <4 is considered normal; a score >7 indicates impairment severe enough to be classified as dementia.
** HAD = HIV-associated dementia; MCMD = Minor Cognitive and Motor Disorder; Subsyndromic = Mildy abnormal test performance without symptoms of impairment.
*** Geographical locations in the United States of America: NYC = New York City; TEX = Galveston/Houston, Texas; UCSD = San Diego, California; UCLA = Los Angeles, California.
**** WT = Caucasian; BL = African American; AS = Asian; HP = Hispanic

SUBJECTS ARRANGED ACCORDING TO SELECTION CRITERIA

GENOMICS LABORATORY SAMPLE IDENTIFICATION CODE

Last two digits: A unique sample identifier used by the UTMB Molecular Genomics Laboratory.

The samples are labeled from 1 through 72.

† Denotes that the isolated RNA had potentially significant hydrolysis when analyzed using gel chromatography, which showed evidence of RNA fragments with reduced size. Some, but not all samples with evidence of hydrolysis eventually exhibited evidence of atypical hybridization patterns on the array. See below.

¶ Denotes that the hybridization signal pattern of this sample was influenced by RNA degradation when it was analyzed. One or more outcomes were used to screen for anomalous hybridization on a particular chip. See 6.0 for details. Briefly, a hybridization signal sometimes was present in a relatively high percentage of all the probe sets; alternatively signal intensities were weak across the board and required a high scaling factor for calibration. Both of those indexes suggest that probe sets that did not produce a significant hybridization signal in most of the samples might erroneously report that the signal is present on these particular chips. Another problem index is when spiked control RNA showed an uneven distribution of cDNA representation, due to uneven conversion to cDNA. This means that the effectiveness of hybridization of some transcripts on the chip, but not all transcripts on the chip, may have been corrupted due to RNA degradation. These errors can produce over- or under-representation of certain signal intensities relative to the other transcripts on the array.

BOLD denotes that these two cases were added to compensate for cases that had preliminary evidence of potentially substantial RNA degradation.