Nanopore sequencing at Mars, Europa, and microgravity conditions

Life as we know it uses nucleic acids as the basis for heredity and information polymerization that utilize biological or informational building blocks, common physicochemical scenarios for life’s origin, or common ancestry via meteoric exogenous, most plausible for Earth and Mars. Beyond the search for life, sequencing is high relevancy supporting human health on Earth and in space, from detecting infectious diseases, to monitoring of biologically-based life support systems. Nanopore sequencing, as commercialized by Oxford Nanopore Technologies, is a promising approach that is now used ubiquitously in the lab and in the field. McIntyre et al. reported a single mapped read obtained via nanopore sequencing during parabolic flight, obtained across multiple parabolas. Vibration of fly boxes revealed that 70% of probes survive launch, while the most successful nanoplore sequencing on the International Space Station (ISS). However, we are not aware of any nanopore experiments that specifically study the impact of vibration while sequencing. Here we test the impacts of: (1) altered g level, (2) vibration, and (3) updated chemistry/fly cells.

Prior work demonstrated feasibility of nanopore sequencing in space-related contexts: (a) single mapped read during parabolic flight, and (b) successful nanopore sequencing aboard the International Space Station (ISS). However, the literature lacked dedicated studies on how in-flight vibration and altered gravity (reduced or hypergravity) affect sequencing performance, read quality, ionic current noise, and base translocation times. This study addresses these gaps by directly measuring sequencing under controlled parabolic flight phases with quantified vibration spectra and g levels.

Experimental setting: Parabolic flight operations were conducted on November 17, 2017 aboard a Boeing 727-200 aircraft (G-Force One, Zero Gravity Corporation). Four sets of parabolas (5, 6, 4, and 8 parabolas) were flown. The first set targeted, in order, Mars g, Mars g, Lunar/Europa g, and 0 g; remaining parabolas targeted 0 g. Flight phases were segmented using accelerometer data into transition, parabola, hypergravity, and other (steady flight/turns). Sequencing: Control lambda DNA was sequenced using an Oxford Nanopore MinION. Sequencing was performed on the ground (38 min) and during flight (103 min) using the same flow cell. Reads were categorized by flight phase for analysis; mux reads were excluded for certain statistical analyses to avoid start-up effects. Basecalling provided per-base quality scores; a representative read quality score q was computed as q = -10*log10(p), where p is the mean read-base error probability. Instrumentation for acceleration/vibration: A triaxial piezoelectric accelerometer (TE Connectivity model 81) sampled at 5 kHz was mounted adjacent to the MinION on a common baseplate using double-sided tape to maintain near-unity frequency response. Earth-relative acceleration magnitude g was computed as sqrt(gx^2 + gy^2 + gz^2). Vibration analysis and filtering: Power spectral density (PSD) was computed via Welch’s method. To focus on frequencies relevant to base translocation (<10 ms timescales) and avoid poor sensor response near DC, a high-pass IIR filter (stopband 5 Hz to 60 dB attenuation, passband 10 Hz with unity ripple, sample rate 5 kHz) was designed (MATLAB designfilt) and applied using zero-phase filtfilt. RMS vibration was computed in 1 s bins (MATLAB rms). PSDs were recomputed post-filtering. Sequencing data processing and alignment: Fastq outputs were adapter-trimmed using Porechop. Untrimmed reads were aligned to the lambda reference genome (NEB Lambda; NCBI NC_001461.1 with noted mutations) using Tombo (docker image tombo2). Tombo associated raw ionic current signals with genomic bases to estimate base-level translocation times and ionic current noise (norm_std). Coverage and alignment statistics were computed, including percentage of bases aligning to the reference. Integration of sequencing and acceleration data: Read timestamps were offset-adjusted to align with accelerometer data. Reads and bases were assigned to flight phases (parabola, transition, hypergravity, other) based on timing overlap. Median read quality and ionic current noise time series were computed in 1 s bins. Statistical analyses: Stepwise linear regression (MATLAB stepwisefit/stepwise) assessed relationships among time, RMS vibration, and g level (flight only) with median read quality and ionic current noise. For flight analyses, recursive elapsed time windows (e.g., 4000 s) were used to reduce confounding from takeoff/landing. One-way ANOVA (anova1) with Tukey’s HSD examined differences in read quality and ionic current noise across flight phases. A two-sample Kolmogorov–Smirnov test (kstest2) compared base translocation time distributions between ground and flight. Coverage across parabolas was modeled versus parabola duration (reporting adjusted R²).

- Sequencing feasibility across g levels: Reads were successfully acquired during all parabolas, including Mars g (~0.378 g), Lunar/Europa g (~0.166 g), and 0 g conditions, demonstrating functional nanopore sequencing under reduced gravity and dynamic flight phases. - Throughput: Using the same flow cell, 38 min ground sequencing yielded 5,293 reads; 103 min in-flight sequencing yielded 13,823 reads. Reads were obtained in all phases, including parabola, hypergravity, transition, and other flight segments. - Vibration environment: Filtered RMS vibration was substantially higher in flight than on ground (average ~5.89× higher). Dominant aircraft-associated vibration peaks were observed in 1 Hz–1 kHz with notable peaks near 116–128, 250–270, 495–496, 580–680, and 876 Hz. Freefall (0 g) segments showed reduced vibration compared to other phases. - Read quality vs. phase: One-way ANOVA with Tukey’s HSD indicated significant differences in read quality among phases; read quality was lowest during parabolas (median qp ≈ 8.3) and highest during hypergravity. - Regression on read quality: On ground, time was the only marginal predictor of sequence quality (p = 0.06) with low explained variance (R² = 0.060). In flight, time, RMS vibration, and g level were statistically significant predictors (all p < 1e-4) but explained variance remained modest (R² = 0.275). - Base alignment and coverage: Tombo-aligned base percentages were high and comparable between conditions (ground 87.8%, flight 89.7%). Average coverage during individual parabolas was sufficient to cover the lambda genome multiple times. Coverage was largely explained by parabola duration (adjusted R² = 0.807). - Translocation times: Despite much higher in-flight vibration, base translocation time distributions were very similar between ground and flight. Median translocation times were identical (1.8 ms); mean times differed slightly (ground 2.2786 ms vs. flight 2.4035 ms). A K–S test detected a statistically significant but small shift toward longer translocation times in flight (est. statistic ≈ 0.0306), indicating robustness of translocation to vibration. - Ionic current noise: Stepwise regression implicated time as a significant predictor of ionic current noise during flight; phase analysis showed ionic current noise was lowest during hypergravity and comparable between parabola and transition phases. Differences among phases were statistically significant but of very small magnitude (largest mean difference ~0.003 in normalized units). - Example mapping: A high-quality “Mars” read (length 6,402 nt) produced top BLAST alignment to Enterobacter phage lambda (J02459.1), with 92% identity over 6,651 bases (gaps 5%). - Zero-g fidelity: Parabolic 0 g segments had residual accelerations (~0.041 ± 0.005 g), still demonstrating sequencing feasibility under very low g comparable to or lower than many target environments except the very low-g Enceladus surface (~0.011 g).

The study directly addresses whether nanopore sequencing performance is affected by altered gravity and aircraft vibration relevant to spaceflight and planetary operations. Successful read acquisition during Mars, Lunar/Europa, and microgravity-equivalent parabolas demonstrates that reduced g does not preclude sequencing, advancing the case for in situ life-detection approaches based on nucleic acids in planetary missions. Quantitative analyses show that while flight introduces significantly higher RMS vibration, key performance metrics—base translocation times, ionic current noise, and read quality—remain largely robust. Translocation times were nearly unchanged in median, with only a slight increase in the mean during flight, and ionic current noise differences across phases, although statistically significant, were minimal in effect size. Read quality showed modest phase-dependent differences, tending to be higher during hypergravity and lower during parabolas, but regression models explained limited variance, indicating that neither vibration nor g level substantially degrades performance under the tested conditions. These findings imply nanopore sequencing can operate concurrently with other activities that may induce vibration (e.g., drilling) and in dynamic mobile environments. The observed robustness supports use for life-detection missions to Mars, the Moon, and potentially Europa (with consideration for radiation constraints), as well as for microbial monitoring in support of human spaceflight.

This work demonstrates that Oxford Nanopore MinION sequencing functions reliably under reduced gravity (including Mars and Lunar/Europa analog g) and in dynamic, vibration-rich environments encountered during parabolic flight. Performance metrics—including read yield, base alignment rates, translocation times, and ionic current noise—were consistent between ground and flight with only small, largely negligible differences attributable to vibration or g level. These results support the feasibility of deploying nanopore sequencing for in situ life detection and microbial monitoring in space missions and other mobile platforms on Earth. Future directions include: (1) extended-duration sequencing under controlled reduced-g conditions to further quantify long-term effects; (2) integrated testing alongside other mission operations that induce vibration; (3) evaluation of hardware and chemistry under relevant radiation environments (e.g., shielding strategies for Europa); and (4) optimization of sample handling and library stability for field-deployable, autonomous planetary applications.

- Zero-g fidelity: Parabolic “0 g” segments had nonzero mean acceleration (~0.041 ± 0.005 g), higher than Enceladus surface gravity (~0.011 g), limiting direct extrapolation to ultralow-g environments. - Library stability: In one experiment, the DNA library was stored for ~72 h before loading, potentially causing degradation and loss of ligated adapters, which could reduce read quality and nanopore loading efficiency. - Time at each g level: Sequencing durations within individual parabolas were brief, limiting statistical power for phase-specific effects. - Sensor/processing constraints: Vibration measurements were high-pass filtered (>10 Hz) to avoid poor sensor response near DC; thus, low-frequency components (<10 Hz) were excluded from analysis, and filtered measures may not capture all relevant mechanical influences. - Technological evolution: Ongoing improvements in sequencing chemistry and software could contribute to time-related trends independent of vibration or g effects, complicating attribution. - Radiation considerations: While recent work suggests MinION components can tolerate radiation doses relevant to several planetary missions, Europa may require additional shielding; this study did not test radiation effects.